Socio-Economic Impact Assessment of Hydrogen Injection in the Natural Gas Network ^†

Abstract

This study explores the feasibility parameters of a potential investment plan for injecting “green” hydrogen into the existing natural gas supply network in Greece. To this end, a preliminary profitability optimization analysis was conducted through key performance indicators such as the cost of hydrogen and the socio-environmental benefit of carbon savings, followed by break-even and sensitivity analyses. The identification of the major impact drivers of the assessment was based on the examination of a set of operational scenarios of varying hydrogen and natural gas flow rates. The results show that high natural gas capacities with a 5% hydrogen content by volume are the optimal case in terms of socio-economic viability, but the overall profitability is too sensitive to hydrogen pricing, rendering it unfeasible without additional motives, measures and pricing strategies. The results feed into the main challenge of implementing commercial “green” hydrogen infrastructures in the market in a sustainable and feasible manner.

1. Introduction

Decarbonization is both a complex and an expensive process, as greenhouse gas (GHG) emissions continue to rise [1] and stakeholders of all levels are trying to find ways to adapt and invest in “greener” alternatives [2,3]. In this context, injecting hydrogen into existing natural gas (NG) networks is a step towards reducing GHG emissions that—if performed successfully- presents some distinct advantages. The first is that no additional costs need to be incurred for the construction of new hydrogen transport infrastructure. Second, blending hydrogen at relatively low concentrations into the gas mix does not require modifications of end use devices (heaters, boilers, etc.) [4]. Third, it can reduce dependency on conventional energy sources and increase the resilience of the gas supply in the case of supply disruptions. Injected hydrogen into NG pipelines can either be sold and used as a hydrogen/NG blend for all usual cases (heating) or separated again from the NG at other downstream plants and sold as pure hydrogen; this paper considers only the former case.

The timing for such initiatives is also rather promising as the European Union has established a favorable regulatory framework around hydrogen use through many supportive policies. These include, among others, the Hydrogen Strategy, the Renewable Energy Directive, the Clean Hydrogen Alliance, the Gas Market Directive, the Hydrogen Bank, the Innovation Fund and the Hydrogen Valleys [5,6,7,8,9,10,11,12]. The hydrogen market is also emerging in several European countries, and Greece is no exception; its updated National Plan for Energy and Climate (NECP) [13] acknowledges the potential of hydrogen use, but also notes that it is still a “non-mature technology”. In this context, the integration of green hydrogen into the energy market (either by itself as a “clean” combustible fuel or by blending it with NG) has been explored, with some good practices already having been documented. Indicative examples include the projects HYFLEXPOWER, HISELECT, FenHYx, HyNTS FutureGrid, HyDeploy, H21 Leeds City Gate, HyREADY, H2Future HRHYD and 24/7 ZEN, where a multitude of EU countries were involved [14,15,16,17,18,19,20,21,22,23]. Stakeholders were gas companies, energy infrastructure providers, research institutions and companies, diverse groups and alliances of related interests, and start-ups. There is also a past work that demonstrated hydrogen injection in the Greek natural gas network, highlighting the key position of Greece as a potential part of a hydrogen export corridor, connecting Europe with the Middle East and North Africa [24,25].

Following on from these developments, the purpose of this study was to investigate the feasibility of injecting hydrogen into the existing NG network in Greece. To this end, a qualitative and quantitative feasibility study was performed, providing estimates for the selling price of both hydrogen and a hydrogen/NG blend. Knowing the price ranges that are viable is a prerequisite for planning any investment in hydrogen technologies; thus, the findings of this paper can contribute to the broader economic assessment of a renewable market around hydrogen and, therefore, the development and operation of large-scale “green” hydrogen production units.

The remainder of this paper is structured as follows: Section 2 contains an overview of the main types, characteristics and costs of hydrogen and information about hydrogen injection systems from the literature and previous use cases. Section 3 presents the methodological framework used. Section 4 contains the results and the ensuing sensitivity analysis, which are then discussed in Section 5. Section 6 concludes this paper with some key remarks and refers to the limitations and future research directions emanating from this study.

2. Theoretical Background

There are many types of hydrogen, each labeled with a different color based on the manner of its production. “Green” hydrogen, which is the focal point of this paper, is produced through water electrolysis, using excess electricity from renewable energy sources that cannot be absorbed by the electricity grid and would otherwise be lost [26]. While the production capacity of “green” hydrogen is currently low due to its high production costs [27], future projections are promising due to the low GHG emissions associated with its production [27]. As the key enabling parameter for the implementation of large-scale “green” hydrogen production is its economic feasibility, hydrogen injection into the NG grid could—at least in theory—boost a gradual transition to “green” hydrogen.

As far as other hydrogen “colors” are concerned, “grey” hydrogen is produced from methane reforming processes [27] and corresponds to the largest fraction of total hydrogen production [27]. “Blue” hydrogen is a low-carbon hydrogen category as it is produced from fossil fuel plants accompanied by a carbon capture and storage (CCS) system [27]. There are also some other less common categories, such as “pink”, “red”, “purple”, “yellow”, “turquoise”, “brown”, “black”, “aqua” and “white” hydrogen [27]. Regardless of the type used, the hydrogen/NG blend is sometimes called “hythane” [28].

Comparisons between the many different hydrogen types are usually based on the average normalized cost and corresponding CO₂ emissions associated with their production: “yellow” hydrogen is the most costly type (~7.4 USD/kg of hydrogen), followed by “green” hydrogen (~4.8 USD/kg) and “purple”/“pink”/“red” hydrogen (~4.0 USD/kg) [27]. The cost of “turquoise” hydrogen is 2.0 USD/kg, while “black”/”brown” and “blue” hydrogens have slightly lower costs (~1.6 USD and ~1.5 USD/kg, respectively) [27]. The least expensive categories are “grey” and “aqua” hydrogens, with 1.0 USD and 0.2 USD/kg. Considering the environmental impact, “black”/”brown” hydrogen is the most impactful category, with 21.5 kg of CO₂-eq/kg of hydrogen [29]. At much lower levels, ”grey” and “turquoise” hydrogens are linked to 10.3 kg and 3.4 kg CO₂-eq/kg of hydrogen, respectively [29]. “Green” hydrogen does not have direct CO₂ emissions; yet, the total GHG emissions are not zero (approximately 1.8 kg CO₂-eq/kg of hydrogen) [27,29]. “Yellow”, “blue” and “purple”/”pink”/”red” hydrogen categories have lower GHG emissions (1.7, 1.6 and 0.3 kg CO₂-eq/kg of hydrogen, respectively) [29]. The cost and GHG emissions of the different “colors” of hydrogen are summarized in

Table 1. Summary of economic and environmental indicators for different hydrogen categories (average values) [27,29].

Hydrogen Type	Cost (USD/kg)	GHG Emissions (kg CO2-eq/kg)
“Green”	4.8	1.8
“Purple”, “pink”, “red”	4.0	0.3
“Yellow”	7.4	1.7
“Grey”	1.0	10.3
“Black”, “brown”	1.6	21.5
“Turquoise”	2.0	3.4
“Blue”	1.5	1.6

.

At the same time, the prices of “green” hydrogen are also dependent on the fluctuations of interest rates and inflation. A study [30] investigating the impact of those two parameters highlighted that the price range of “green” hydrogen is EUR 3.47–EUR 5.46/kg, while the levelized cost of energy hydrogen (OPEX plus CAPEX divided by the energy flow) is EUR 0.104–EUR 0.164/kWh, for an interest rate range between 4% and 10% and considering inflation rates of 2%, 4% and 6% [30].

A similar assessment [31] of hydrogen injection into the NG grid took into account a price between USD 2 and 22/MBTU of NG. The maximum hydrogen content used is 6% by volume, while a 10% content can be allowable but under specific circumstances. The scenarios taken into account were for 5%, 10% and 20% of hydrogen volumetric content. Since hydrogen has a lower calorific value than NG, the hydrogen/NG blend had a gross calorific value of 36.5 MJ/Nm³ with 20% hydrogen (which amounted to a 13% decrease from the assumed NG gross calorific value of 42.1 MJ/Nm³). The “green” hydrogen technologies of alkaline electrolyzer, polymer electrolyte membrane and solid oxide electrolyzer were preferred in this assessment [31] against the “blue” hydrogen technologies of methane reforming combined with carbon capture. The capital expenditure was linearly correlated with the hydrogen content: EUR 112.83M for 5% hydrogen, EUR 225.65M for 10% hydrogen and EUR 451.31M for 20% hydrogen.

Regarding the costs of transporting the hydrogen, roughly EUR 0.57M per 39.76 km of pipeline (or EUR 14.3k per km) is needed for the refurbishment of existing NG pipelines, while EUR 1.50M per 0.750 km (or EUR 2.0M per km) is needed for new hydrogen pipelines [31]. Moreover, the hydrogen needs to be compressed before it is injected into the grid, adding an additional EUR 0.65M per 1500 MWh (or EUR 0.43M/MWh) for compressors. The expected net cash flow in [31] was expected to be positive after the eighth year of the investment, while the carbon dioxide emission reduction was roughly half of the hydrogen percentage in the NG mix. More specifically, the emissions were expected to be reduced by 2.56% for the 5% hydrogen scenario, while the reduction was estimated to be 5.26% and 11.11% for the 10% and 20% hydrogen scenarios, respectively [31].

Another study [28] investigated 576 scenarios for 2026–2050 to determine the GHG mitigation potential and related cost-effectiveness, considering several carbon pricing policies (0–350 USD/ton of carbon dioxide). The hydrogen content range in the study was 1–15% by volume. The expected price of blue hydrogen was over 13 USD/GJ. Since the price of conventional NG is 2–4 USD/GJ, an energy transition to a “hythane” blend may challenge the social acceptance of the new fuel, while it is also challenging for pure NG-dependent consumers [28].

Another study [32] included both hydrogen production and injection processes regarding the control volume of its economic assessment. Using a 5.86% hydrogen blend by volume, the production equipment cost was 1259.26 USD/Nm³/h, while the compressor cost was 296.30 USD/Nm³/h. The main cost drivers for the capital expenditure were the electrolyzer and the compressor (71.27% and 16.77% of the total capital expenditure), while the main cost drivers for the operational expenditures were those related to electricity and transportation (55.21% and 34.68% of operational expenditure). The cost of the blend was estimated to be 0.244 USD per Nm³ and the cost of pure hydrogen was estimated to be 0.303 USD/Nm³. A sensitivity analysis indicated that the electricity and transportation costs had higher sensitivity rates. The estimated payback period was rather long, 13–18 years, with an estimated internal rate of return (IRR) of approximately 16.26% [32].

An assessment [33] of hydrogen production from NG reforming considered an NG blend with 95% mol methane to produce 18.9 tons of hydrogen per hour. The price of NG was 2.65 USD/GJ and the price of electricity was 0.108 USD/kWh. The hydrogen production unit was estimated to operate 365 days per year and the assessment considered a 30-year horizon for the investment with a 12% depreciation rate. The capital expenditure took into account the NG reforming, water–gas shift reaction, carbon dioxide removal, boiler, steam turbines, pressure swing adsorption and carbon dioxide liquefaction processes, depending on several different cases (reformer types and “color” of produced hydrogen: grey without carbon capture, blue with carbon capture), while the operational expenditure included labor, utilities, maintenance and repair, operating supplies, laboratory charges, plant overheads, administrative costs, supply chain functions, and research and development activities.

The key performance indicators for the economic impact assessment were the thermal efficiency and the minimum hydrogen selling price, which was identified by setting the net present value (NPV) of the overall investment plan to be equal to zero. Depending on the case, the minimum selling prices were 0.98–1.07 USD/kg of “grey” hydrogen and 1.11–2.16 USD per kg of “blue” hydrogen. A sensitivity analysis showed the impact of the NG price on the results, which is a timely geopolitical issue. More specifically, for an NG selling price of 0.5 USD per GJ, the minimum selling prices were 0.4–0.5 USD/kg of “grey” hydrogen and 0.5–1.3 USD/kg of “blue” hydrogen, while, for an NG selling price of 13.5 USD per GJ, the minimum selling prices were 3.8–5.7 USD/kg of “grey” hydrogen and 3.9–6.7 USD per kg of “blue” hydrogen. In an additional environmental impact assessment implemented by the same study, the carbon dioxide emissions were estimated to be 8.7–11.7 kg/kg of produced “grey” hydrogen and (−0.9)–2.3 kg/kg of “blue” hydrogen. The carbon dioxide emissions were monetized using a range of carbon prices (0–100 USD per t of carbon dioxide) [33].

In another assessment [34], a photovoltaic–hydrogen power system was benchmarked against an existing photovoltaic–diesel power system, for a maximum daily peak power demand of 8.3 kW. For the existing system, 65% of the total 24.1 MWh production was obtained from solar energy, while the remaining 35% was from diesel. The estimated carbon dioxide emissions were 8764 kg/y, while the total annualized cost was 20.7k EUR per year: 11.4k EUR/year for capital expenditure, 5.7k EUR/year for the replacement cost, 1.0k EUR/year for operation and maintenance, and 2.7 EUR/year for the fuel. The photovoltaic–hydrogen power system was estimated to produce 39.3 MWh, which was 100% renewable (72% solar energy, 28% hydrogen fuel cell). There were no carbon dioxide emissions in this case, while the total annualized cost was 31.4k EUR per year (approximately 1.5 times greater than that of the photovoltaic–diesel system): 28.8k EUR per year for the capital expenditure, 1.8k EUR per year for the replacement cost, 0.9k EUR per year for operation and maintenance, and 0 EUR per year for the fuel (since no diesel generator was needed in this case) [34].

Hydrogen can also be produced from sources generally regarded as waste. An assessment of this possibility [35] estimated the levelized cost of hydrogen (OPEX plus CAPEX divided by the hydrogen mass flow rate), the internal rate of return and the return on investment using several feedstocks, which were categorized into two methods: dark fermentation and gasification. For the dark fermentation processes, the levelized cost of hydrogen was 1.02–3.20 USD/kg, while the internal rate of return was 9.25–24.07% and the return on investment was 12.8–60% [35]. For the gasification processes, the levelized cost of hydrogen is 12.75 USD/kg, while the internal rate of return was 8.4–17.1% and the return on investment was 10.5–16.5% [35].

3. Methodology

The methodological approach of this study consisted of multiple stages. It began with a SWOT analysis, which served as a qualitative assessment of the investment. Then, to estimate the quantitative impact, the control volume of the system was determined, and the mass and energy balances of the investment were estimated. These fed the cost analysis with the terms of capital and operational expenditures (CAPEX and OPEX), and then the potential revenue from selling the hydrogen was estimated. The minimum selling price for hydrogen was estimated through a break-even analysis. The results were then optimized on the basis of some key performance indicators (KPIs). Lastly, a sensitivity analysis was performed in order to assess the impact of fluctuations of some crucial parameters on the investment’s feasibility. All methodological stages are shown below in Figure 1.

As discussed in the introduction, the actual hydrogen production was not taken into account in this study (Figure 2). Instead, the control volume began from the point of hydrogen injection into the NG grid (i.e., where the hydrogen is sold and supplied from the producer to the distributor) and ended with the sale and supply of hydrogen from the distributor to the end-users. Firstly, the hydrogen distribution through the grid was investigated in order to estimate the selling price of the fuel, which was then evaluated from a socio-economic perspective.

As for the mass balances of the investment, five NG flow cases were taken into account (

Table 2. NG flow scenarios.

Case	Capacity	NG Flow (m3/h)
1	5%	264.6
2	25%	1323
3 (Base case)	50%	2646
4	75%	3960
5	100%	5292

). Each one in turn contained three sub-cases of hydrogen content (

Table 3. Hydrogen flow scenarios. Base-case scenario: case 3, 5% flow.

NG Flow			Mass Hydrogen Flow (t/h)			Volumetric Hydrogen Flow (m3/h)
Case	Mass Flow (t/h)	Volumetric Flow (m3/h)	1%	5%	10%	1%	5%	10%
1	21	264.6	0.019	0.095	0.19	2.646	13.23	26.46
2	105	1323	0.095	0.475	0.95	13.23	66.15	132.3
3	210	2646	0.19	0.95	1.9	26.46	132.3	264.6
4	315	3960	0.285	1.425	2.85	39.6	198	396
5	420	5292	0.38	1.9	3.8	52.92	264.6	529.2

). This resulted in fifteen (5-by-3) total cases of hydrogen being included in this study, as shown in Table 3 and Figure 3. Of these, the middle scenario (3) was chosen as the base case comprising 5% hydrogen flow, as it was both a convenient middle point for comparisons and used a similar concentration to many of the aforementioned studies.

There are several empirical methods for the estimation of capital costs; however, the error margin was significant, and the estimations could vary from EUR 2.5M to EUR 50M. In this study, the costs for the base case scenario were estimated to be EUR 20M (Figure 4) for the hydrogen compression unit and injection facility (a 1.5 MW compression and injection unit was considered), inclusive of various other costs and contingency funds. Nevertheless, this value was further investigated through several scenarios, described below.

Since a reference point (base case capacity and cost) was known, the investment cost could be estimated for all cases through scaling up or scaling down, based on a rule of thumb (Equation (1)) [36]:

C = C_ref ∙ (Q/Q_ref)^e

(1)

where

• C is the investment capital expenditure;
• C_ref is the reference capital expenditure;
• Q is the capacity of the investment in NG;
• Q_ref is the reference capacity;
• e = 2/3, which is a regularly used scaling exponent for compressors.

As for the OPEX, it covered both the electricity costs (base case estimation: 75 EUR/MWh) and the maintenance costs (6% of CAPEX annually [37]). The cost breakdown of all 5 cases is shown below in

Table 4. Cost breakdown for all cases. Base case scenario: case 3.

Case	Capital Cost (kEUR)	Electricity Cost (kEUR/y)	Maintenance Cost (kEUR/y)
1	4272	89	256
2	12,558	446	753
3	19,980	891	1199
4	26,217	1337	1337
5	31,790	1782	1782

. It is important to clarify that the hydrogen production plant was not included in the boundary limits of this study and, therefore, the electricity cost was not expected to be a major economic impact driver.

It should be noted at this point that the uncertainty regarding capital cost estimation was quite significant. The costs of the compression and injection facilities could be different, while the state of the existing gas distribution pipelines and equipment could necessitate some upgrades and modifications. To account for this, some additional scenarios for the capital cost estimation were considered (

Table 5. CAPEX scenarios. Base case scenario: case 3, tier 2.

Case	Tier 1 (kEUR)	Tier 2 (kEUR)	Tier 3 (kEUR)	Tier 4 (kEUR)
1	2136	4272	10,679	21,358
2	6279	12,558	31,394	62,788
3	9990	19,980	49,950	99,900
4	13,108	26,217	65,541	131,083
5	15,895	31,790	79,474	158,948

). The cumulative effects of each scenario tier on total CAPEX are shown in Figure 5.

According to the hydrogen policy of the European Union, the goal for the green hydrogen selling price is about 2.5–5.5 EUR/kg [38], but, in practice, prices tend to be higher due to high production costs (electrolysis, compression, transportation, etc.). It is obvious that this range is significant (more than 100% difference); thus, calculations were made with a base case scenario of the average price (4 EUR/kg), which was equivalent to 0.096 EUR/kWh since the calorific value of hydrogen is approximately 150 MJ/kg. This cost was also considered for the hydrogen supply from the electrolysis unit to the distributor (this paper’s control volume). As the scope of this study extended from the point of injection of hydrogen into the gas network until the sale of the gas mix to end-users, the production stage of hydrogen was not included in the cost analysis.

Since the hydrogen flow in the 15 cases was known, the hydrogen supply cost (hydrogen sale from producer to distributor) could be estimated for each one. The results are shown in

Table 6. Hydrogen supply cost scenarios. Base case scenario: case 3.

Case	1% Hydrogen Cost (kEUR/y)	5% Hydrogen Cost (kEUR/y)	10% Hydrogen Cost (kEUR/y)
1	602	3010	6019
2	3010	15,048	30,096
3	6019	30,096	60,192
4	9029	45,144	90,288
5	12,038	60,192	120,384

.

For the sales revenue estimation (hydrogen sale from distributor to end-users), the hydrogen pricing model was based on the existing NG pricing model in Greece (Equation (2)) [39]:

C = x ∙ C_Q ∙ Q + C_E ∙ E

(2)

where

• C is the cost of energy (EUR/kWh);
• C_Q is the capacity charge rate (EUR/(kWh/h)), which was considered equal to the weighted average rates for NG per region and end-use in Greece (0.68996 EUR/(kWh/h) [40,41,42,43]);
• Q is the total reserved capacity from end-users in Greece (42,329 MWh/h [42,43]);
• C_E is the energy consumption rate (EUR/kWh);
• E is the consumed energy (kWh);
• x is the hydrogen energy contribution to the total NG consumption (56.64 TWh/y [44]).

Solving Equation (2) for the hydrogen energy contribution parameter, the reserved capacity revenue could be calculated (shown in

Table 7. Hydrogen reserved capacity revenue calculation. Base case scenario: case 3 (5% hydrogen).

	Case	1% Hydrogen	5% Hydrogen	10% Hydrogen
Hydrogen energy flow (TWh/y)	1	0.00627	0.0314	0.0627
	2	0.0314	0.157	0.314
	3	0.0627	0.314	0.627
	4	0.0941	0.470	0.941
	5	0.1250	0.627	1.250
Hydrogen contribution to annual consumption	1	0.01%	0.06%	0.11%
	2	0.06%	0.28%	0.55%
	3	0.11%	0.55%	1.11%
	4	0.17%	0.83%	1.66%
	5	0.22%	1.11%	2.21%
Reserved capacity revenue (kEUR/y)	1	3.2	16	32
	2	16	81	162
	3	32	162	323
	4	49	243	485
	5	65	323	647

).

The energy consumption rate of hydrogen was determined through a break-even analysis (solution of Equation (3)) for each of the 15 hydrogen flow cases:

Revenue − Total Annualized Cost = 0

(3)

which was equivalent to Equation (4):

P_hydr = (C_hydr,supply + C_elec + C_maint + e ∙ CAPEX − R_capac) / E_hydr

(4)

where

• P_hydr is the selling price of hydrogen to the end-users;
• C_hydr,supply is the annual hydrogen supply cost;
• C_elec is the annual cost of electricity;
• C_maint is the fixed annual maintenance cost;
• e is the CAPEX annualization factor (0.1, if 10-year horizon is assumed);
• CAPEX is the investment capital expenditure;
• R_capac is the fixed reserved capacity revenue;
• E_hydr is the hydrogen selling quantity to the end-user.

A key performance indicator (KPI) for the socio-economic impact assessment was the increase in fuel prices compared to conventional NG fuel. The blend price was calculated as follows (Equation (5)):

P_blend = y_hydr ∙ P_hydr + y_ng ∙ P_ng

(5)

where

• P_blend is the selling price of the hydrogen/NG blend fuel to the end-users;
• y_hydr and y_ng are the mass fractions of hydrogen and NG in the blend;
• P_hydr is the selling price of hydrogen to the end-users (Equation (4));
• P_ng is the average selling price of NG to the end-users in Greece (0.00512 EUR/kWh [40,41]).

Then, the fuel price increase could be calculated using Equation (6):

%ΔP = P_blend/P_ng − 1

(6)

where

• %ΔP is the fuel price increase;
• P_blend is the selling price of the hydrogen/NG blend fuel to the end-users (Equation (5));
• P_ng is the selling price of NG to the end-users.

The KPI used for capturing the socio-economic benefits of the investment was the decrease in carbon dioxide emissions. As the mass flows of methane (the main component of NG) and carbon dioxide are linearly dependent, the CO₂ emissions ratio of the blend fuel to NG was equal to the mass fraction of methane (or NG) in the blend. The reduction of emissions was the difference of this ratio from 1 (or 100%), as shown in

Table 8. Carbon dioxide savings estimation. Base case scenario: case 3, 5% hydrogen.

	Case	0% Hydrogen	1% Hydrogen	5% Hydrogen	10% Hydrogen
Carbon dioxide emissions (kt/y)	1	457	457	455	453
	2	2287	2285	2277	2266
	3	4574	4569	4553	4533
	4	6861	6855	6829	6799
	5	9148	9139	9106	9066
Carbon dioxide emission savings (kt/y)	1	0	0.41	2.06	4.10
	2	0	2.07	10.30	20.51
	3	0	4.13	20.60	41.01
	4	0	6.20	30.90	61.52
	5	0	8.27	41.20	82.02
Carbon dioxide emission savings	1	0	0.1%	0.5%	0.9%
	2	0	0.1%	0.5%	0.9%
	3	0	0.1%	0.5%	0.9%
	4	0	0.1%	0.5%	0.9%
	5	0	0.1%	0.5%	0.9%
Monetized carbon dioxide emission savings (kEUR/y)	1	0	64	319	635
	2	0	320	1595	3175
	3	0	640	3190	6351
	4	0	960	4784	9526
	5	0	1280	6379	12,701

. The savings could be monetized with a base case social cost of CO₂, which is generally given as about 155 EUR/tonCO₂ [45].

Due to the many degrees of freedom present in the input variables, a sensitivity analysis was performed at the last stage of the methodological approach to assess how the values of the KPIs we used were affected as a result of changes in specific input variables relative to the base case scenario. The results were benchmarked against each other (using unitless relative changes) in order to distinguish the most sensitive input parameters.

4. Results

4.1. SWOT Analysis

First, a SWOT (Strengths–Weaknesses–Opportunities–Threats) analysis was implemented to offer a qualitative overview of the key investment points, as shown in Figure 6. The major strength of hydrogen injection into the NG network was the partial mitigation of the environmental and social impact of using the gas mix without the need to design, buy and construct new infrastructure. At the same time, using a hydrogen/NG blend fuel offered a new energy solution, extending the energy portfolio. However, ensuring a constant hydrogen demand can be quite challenging. Furthermore, hydrogen’s thermodynamic properties are significantly different than those of NG, necessitating a thorough thermodynamic study before implementation in specific (sensitive) use cases.

The general unavailability of market data was expected to hinder revenue and cost estimations. On the other hand, the potential of hydrogen injection into the NG network seems very promising as the EU supports hydrogen use and has established supporting policies within a favorable regulatory framework. The scale of hydrogen production and its share within gas mixes is expected to gradually increase, paving the way to a hydrogen economy, aligned with the EU’s climate and energy framework. Through this effort, the knowledge-intensive parts of the hydrogen ecosystem are expected to push developments in technology and innovation, with a considerable number of new jobs being created in the process. Nevertheless, hydrogen’s properties and flammability will need to be thoroughly studied, as will the tolerance of the existing infrastructure to hydrogen. At the same time, the end price increase of the blend should still make sense financially for the final users, which might be hard (or even impossible) to ensure.

4.2. Impact Drivers

Regarding the cost drivers of the investment, hydrogen supply was the most cost-intensive constituent (Figure 7). The total annualized costs could run well over EUR 120M per year, depending on the case.

4.3. Hydrogen Selling Price

Regarding the minimum selling price of hydrogen to the end-users that still rendered the investment viable, larger hydrogen concentrations had the potential for larger profits, and, as a result, the price could be lower. However, the relationship between the break-even price and hydrogen concentration was not linear; more specifically, the ratio between the prices for 1% and 5% concentrations was significantly higher than the ratio between the prices for 5% and 10%, as shown below in Figure 8. This could be attributed to the non-linear correlation of CAPEX and hydrogen capacity (Equation (1)).

4.4. Fuel Selling Price Increase

In contrast to Figure 8, higher hydrogen shares of the gas mix resulted in greater economic impacts on end-users, while fuel prices disproportionately increased (shown below in Figure 9). Taking into account the results shown in Figure 7 and Figure 8, the optimal solution among the three investigated scenarios of hydrogen content was injection at the 5% level. However, with a mix of only 5% hydrogen, the blend fuel’s selling price was approximately two times that of NG.

4.5. Cost–Benefit Analysis

One widely utilized way to benchmark the environmental and social benefits against the financial costs was to monetize the carbon dioxide emission savings [45]. This benefit barely changed among the 5 cases; however, with 5% or 10% hydrogen, the benefit reached approximately 10% of the total annualized cost (Figure 10), which is very promising, especially if a potential increase in hydrogen content in the future is considered.

4.6. Benchmarking

The hydrogen cost in the base case scenario (case 3; 5% hydrogen) was 4.52 EUR/kg. Compared with the corresponding values of green hydrogen from the literature (Figure 11), this selling price was on the same order of magnitude as cases from the literature. More specifically, the selling price was very close to that of [30], hinting that “green” hydrogen markets tend to be formed in similar price ranges close to 4.5 EUR/kg. In comparison to the base case scenario of this study, the selling price could be further decreased and be even more competitive for larger NG flow or greater hydrogen content, as shown in Figure 8. One study [46] took into account a worst-case scenario of 12 EUR/kg of hydrogen, which caused asymmetry to the price curve that was taken into account and, thus, significantly increased the median value. On the other hand, the selling price of [32] was based on 2020 data and as a result did not take into account the price increase and the global energy crisis from 2022 onwards.

4.7. Sensitivity Analysis

4.7.1. Capital Cost

Apart from the base case scenario (EUR 20M for case 3), additional total capital expenditures of EUR 10M, EUR 50M and EUR 100M were inserted into the analysis, as shown in Figure 5. This pushed the fuel selling price upwards (as shown in paragraph 3.4) from 96% to 141% (Figure 12) and was used to account for the worst-case scenario of expensive investments being needed to upgrade the existing infrastructure.

4.7.2. Electricity Cost

The cost of electricity used by the compression and injection facility was 50, 100 and 200 EUR/MWh in three scenarios and 75 EUR/MWh in the base case scenario, as shown in Figure 13. The electricity cost seemed to be a rather insignificant parameter in the sensitivity analysis, confirming the results shown in Figure 7.

4.7.3. Maintenance Coefficient

Similar to the electricity cost, the maintenance cost was also an insignificant cost driver, as shown in Figure 7. As shown in Figure 14, in the base case scenario, the maintenance cost was estimated to be 6% of the capital cost, and two more cases were investigated, where the maintenance coefficient was 2% and 10%.

4.7.4. Hydrogen Supply Cost

As discussed in paragraph 2, the supply cost of green hydrogen (from producer to distributor) needed to be between 2.5 and 5.5 EUR/kg. In the base case scenario, the average value (4 EUR/kg) was taken into account. In the sensitivity analysis, the two boundary values were also considered. According to Figure 15, the impact of this parameter was significant, which confirmed that the hydrogen supply cost was the main cost driver, as shown in Figure 7.

4.7.5. Operating Time

In the base case scenario, an operating time of 330 days per year was taken into account. In the sensitivity analysis, operating times of 300 and 360 days per year were also considered. The difference in the results seemed to be very minor, as shown in Figure 16.

4.7.6. Summary

Summarizing all the above findings, the hydrogen supply cost and the capital expenditure emerged as the most significant cost parameters. The percentage change of blend fuel price compared to the base case scenario (regarding the percentage change of the input variables compared to the base case scenario) was divided into the most significant aspects (hydrogen supply cost, CAPEX) and the least significant ones (electricity cost, maintenance cost, operating time). A summary of the results from the sensitivity analysis is shown below in

Table 9. Sensitivity analysis results.

Category	Parameter	Sensitivity
Raw material supply	Hydrogen supply cost (EUR/kg)	Very high
Cost	CAPEX (EUR)	High
	Electricity cost (EUR/kWh)	Low
	Maintenance cost (% of CAPEX)	Low
Unit functionality	Operating time (days/year)	Low

and Figure 17 and Figure 18.

5. Discussion

This study considered five cases of NG flow, taking into account three scenarios of hydrogen content in the gas blend: 1%, 5% and 10% of volumetric hydrogen content in the NG. The total annualized cost (TAC) of the investment included the hydrogen supply from the producers, the electricity consumption for hydrogen compression and injection, the equipment maintenance cost and the annualized capital cost (for the hydrogen compression and injection equipment). For 1% hydrogen, the TAC was between EUR 1M and EUR 19M; for 5% hydrogen, it was between EUR 4M and EUR 67M; while, for 10% hydrogen, the cost lays between EUR 7M and EUR 127M. These significant differences could be attributed to the high variability of NG flow (265–5292 m³/h) in the five scenarios. The hydrogen supply cost accounted for 45–95% of TAC, followed by the capital cost, which was also a major cost driver. Additionally, the monetized carbon dioxide emission savings from burning less methane at the end uses of the gas mix corresponded to 5–7% of TAC for 1% hydrogen, 8–10% of TAC for 5% hydrogen and 9–10% of TAC for 10% hydrogen.

As far as the selling price was concerned, the minimum for “green” hydrogen was 6.26–9.15 EUR/kg for 1% hydrogen, 4.43–5.01 EUR/kg for 5% hydrogen and 4.21–4.50 EUR/kg for 10% hydrogen. The prices of the latter two scenarios lay in the 2.5–5.5 EUR/kg range, which is established by the European Union and seems to be feasible [38]. However, if the selling price of the fuel blend (i.e., hydrogen/NG blend) was compared to the average NG selling price, the fuel would be approximately 1.5 times more expensive (28–42% price increase) for 1% hydrogen, 2 times more expensive (99-112% price increase) for 5% hydrogen and 3 times more expensive (187–201% price increase) for 10% hydrogen. The latter scenario is not currently feasible. Generally, the hydrogen selling price estimated in this study lay in the same order of magnitude as other similar feasibility studies from the literature.

Lastly, based on the results from the sensitivity analysis (which covered the worst- and best-case scenarios for the design variables), the capital cost and hydrogen supply cost emerged as the most crucial parameters for the investment. More specifically, the fuel selling price increase could vary from 96% to 141% depending on the capital cost scenarios (showcasing the worst-case scenario, if substantial capital needs to be invested in the upgrading of the existing NG grid infrastructure), while it could vary from 66% to 136% depending on the hydrogen supply cost scenarios (showcasing the high dependency on the hydrogen production cost throughout the investment’s time horizon). On the other hand, the cost of electricity for compression and injection, the maintenance cost and the operating time were not significant cost parameters.

The main key performance indicators (KPIs) that were considered for the socio-economic impact assessment are shown below in

Table 10. Results overview. Three volumetric ratios of hydrogen to NG were considered. The results range was attributed to the NG flow range that was considered within the five cases of this study.

Hydrogen/NG Ratio	1%	5%	10%
TAC (million EUR/year)	1−19	4−67	7−127
Hydrogen selling price (EUR/kg)	6.26−9.15	4.43−5.01	4.21−4.50
Fuel selling price increase	28−42%	99−112%	187−201%
Socio-environmental benefit/TAC ratio	5−7%	8−10%	9−10%

.

6. Conclusions

This study implemented a preliminary, multi-staged feasibility analysis of a potential investment plan concerning the injection of hydrogen into Greece’s existing natural gas (NG) supply network. The results indicate that despite the promising potential in terms of mitigating some environmental and social impacts of NG consumption, such an investment plan is still not cost-effective, as the optimal pricing scenarios for “green” hydrogen injection are not economically feasible. Given this, and in line with the overall shift in EU regulations concerning hydrogen adoption, additional measures and motives (e.g., in the form of subsidies or tax exemptions) need to be introduced to render “green” hydrogen/NG mixes a viable and sustainable option for final consumers. Higher hydrogen contents in the country’s fuel mix present significant potential for scaling up several socio-environmental benefits and sustainability while ensuring cleaner fuels and lower GHG emissions.

At the same time, this study is not free of limitations, some of them paving the way for future research. First, the control volume was restricted within the boundaries of the Greek NG grid (from injection to distribution). It can be expanded in future studies by including in the control volume the hydrogen production plant, which will result in a broader range of parameters being considered in the assessment. The capital cost of the injection facility is also restricted to a wide range of scenarios. A detailed design of all the process steps (especially if hydrogen production is included in the analysis), using known cost models, can also provide another future direction. Furthermore, this study investigated a particular pricing and investment scheme, without considering the environmental conditions. Additionally, while some qualitative aspects of the Greek energy system were considered in the analysis, they were not explicitly modeled in the cost structure; this provides an interesting research direction to pursue, but the data availability is expected to be an obstacle. More business models can also be investigated since the current feasibility analysis was highly sensitive to the hydrogen pricing strategy. Lastly, other measurable KPIs can be used for the social impact assessment of “green” hydrogen in conjunction with life-cycle analysis (LCA) to further investigate the potential benefits and externalities of such investments.