Remote Renewable Energy Hubs

• Remote Renewable Energy Hubs (RREHs) are systems integrating renewables, conversion technologies, and logistics to transform remote resources into transportable energy carriers.
• The literature categorizes RREHs by diverse architectures—from isolated microgrids supplying local needs to large-scale export chains producing synthetic fuels.
• Optimization, advanced monitoring, and hierarchical control are key to RREH design, enabling cost-effective operation under regulatory and technical constraints.

Remote Renewable Energy Hubs (RREHs) are renewable energy hubs located in remote areas. In the formal taxonomy, an energy hub “integrates input and output of commodities, conversion, and storage functionalities, enabling coupling between different energy systems,” a renewable energy hub is “an energy hub that relies on renewable energy sources for energy production,” and a Remote Renewable Energy Hub is “a renewable energy hub located in a remote area” (Dachet et al., 10 Jul 2025). Across the literature, the term covers at least two closely related system classes: large export-oriented supply chains that convert high-quality remote renewable electricity into transportable “energy molecules” for distant load centres, and isolated or weak-grid multi-resource microgrids that use local renewables and storage to supply remote communities, offshore platforms, or clustered users (Berger et al., 2021, Toms et al., 8 May 2025).

1. Definition, scope, and taxonomy

The taxonomy paper formalizes an individual RREH as

$$r = (\mathcal{L}_r, \mathcal{G}_r, \mathcal{C}_r, \mathcal{E}_r, \mathcal{I}_r, \mathcal{B}_r, \mathcal{O}_r),$$

where $\mathcal{L}_r$ is the set of locations, $\mathcal{G}_r = (\mathcal{T}_r,\mathcal{H}_r)$ is the technological graph, $\mathcal{C}_r$ is the commodity set, $\mathcal{E}_r$ the exports, $\mathcal{I}_r$ the imports, $\mathcal{B}_r$ the byproducts, and $\mathcal{O}_r$ the local opportunities (Dachet et al., 10 Jul 2025). In this formulation, “remote” is intentionally left subjective; what matters is that the hub is far from major population and demand centres and therefore must export its output via some transport infrastructure (Dachet et al., 10 Jul 2025).

This taxonomy is useful because RREHs are not a single plant type. They can differ by location class, resource mix, conversion chain, export molecule, import dependence, and degree of local integration. An Algerian methane hub, an Algerian ammonia hub, a Greenland hydrogen hub, and an Australian methanol hub can all be RREHs while having different $\mathcal{G}_r$, $\mathcal{E}_r$, $\mathcal{L}_r$0, and $\mathcal{L}_r$1 (Dachet et al., 10 Jul 2025). The paper’s comparison of Algerian CH$\mathcal{L}_r$2 and NH$\mathcal{L}_r$3 hubs makes this explicit: the locations are identical, but the technological graph, commodity set, export set, and byproduct set differ because one hub uses DAC and methanation while the other uses an ASU and Haber–Bosch (Dachet et al., 10 Jul 2025).

From an electrical-system perspective, RREHs overlap with the “Electrical Energy Hub” concept. An Electrical Energy Hub is “a multifunctional node in a power system, managed separately from the main existing control areas, which aggregates local GW-size electrical generation capacity,” and is “characterized by a high density of electrical equipment” with “multiple interconnections to other control areas” (Bastianel et al., 8 Apr 2025). This suggests that offshore, multi-GW, HVDC-interconnected renewable islands are a specific electrical subclass of RREHs. Conversely, not every offshore collector or interconnector node qualifies: the definition requires local multi-GW aggregation and multiple interconnections at the same node (Bastianel et al., 8 Apr 2025).

2. Physical architectures and conversion chains

The canonical export-oriented RREH architecture is a spatially split supply chain. In the North Africa–Europe methane case, the hub has an inland generation cluster with solar PV, wind, battery storage, and an HVDC export line; a coastal processing hub with HVDC reception, desalination, PEM electrolysis, Direct Air Capture (DAC), methanation, H$\mathcal{L}_r$4 storage, CO$\mathcal{L}_r$5 storage, LNG liquefaction, LNG storage, and loading; and a destination terminal with LNG storage, regasification, and gas-grid injection (Berger et al., 2021). The Morocco–Belgium case uses a similar three-module structure: a power production hub in the Western Sahara Atlantic coast desert area, an e-methane production hub near Safi, and an LNG/e-methane hub in Zeebrugge (Fonder et al., 2023).

The conversion chains are commodity-explicit. Primary renewables produce electricity through $\mathcal{L}_r$6 and $\mathcal{L}_r$7. Electrolysis is written as

$\mathcal{L}_r$8

with water commonly supplied by desalination,

$\mathcal{L}_r$9

Carbon- and nitrogen-based fuels then follow different chains. Methanation is represented as

$\mathcal{G}_r = (\mathcal{T}_r,\mathcal{H}_r)$0

DAC as

$\mathcal{G}_r = (\mathcal{T}_r,\mathcal{H}_r)$1

air separation as

$\mathcal{G}_r = (\mathcal{T}_r,\mathcal{H}_r)$2

and Haber–Bosch as

$\mathcal{G}_r = (\mathcal{T}_r,\mathcal{H}_r)$3

(Dachet et al., 10 Jul 2025). In the methane and methanol studies, the synthesis reactions are written explicitly as

$\mathcal{G}_r = (\mathcal{T}_r,\mathcal{H}_r)$4

and

$\mathcal{G}_r = (\mathcal{T}_r,\mathcal{H}_r)$5

(Dachet et al., 2023, Larbanois et al., 10 Jul 2025).

Transport and conditioning strongly shape architecture. CH$\mathcal{G}_r = (\mathcal{T}_r,\mathcal{H}_r)$6 typically requires liquefaction, storage, carriers, and regasification; H$\mathcal{G}_r = (\mathcal{T}_r,\mathcal{H}_r)$7 may require liquefaction and regasification; NH$\mathcal{G}_r = (\mathcal{T}_r,\mathcal{H}_r)$8 is synthesized as liquid at high pressure and temperature and shipped as a liquid; methanol is stored and shipped as an ambient liquid (Fonder et al., 2023, Larbanois et al., 10 Jul 2025). The energy chain is therefore not only generation-plus-synthesis, but a full sequence of production, conversion, storage, and logistics.

A second architectural family interprets RREHs as isolated renewable microgrids. In the offshore microgrid paper, RREHs are, in essence, what that study calls Offshore Renewable Energy Microgrids: islanded systems combining offshore wind turbines, wave energy converters, tidal energy converters, floating photovoltaic systems, and a battery energy storage system to supply offshore platforms or coastal communities (Toms et al., 8 May 2025). This suggests that RREHs are not limited to export-fuel supply chains; they also include stand-alone multi-resource hubs whose “export” function is local service continuity rather than distant molecular transport.

3. Optimization, planning, and regulatory formulations

A defining feature of RREH research is system-level optimization. The graph-based studies formulate the hub as a set of nodes and hyperedges, where nodes represent technologies or subsystems and hyperedges enforce conservation and coupling constraints. In the synthetic methane planning framework, the global problem is written over nodes $\mathcal{G}_r = (\mathcal{T}_r,\mathcal{H}_r)$9 and hyperedges $\mathcal{C}_r$0, with local variables $\mathcal{C}_r$1 and $\mathcal{C}_r$2, and a total objective

$\mathcal{C}_r$3

subject to node-level equalities and inequalities and hyperedge-level equalities and inequalities (Berger et al., 2021). The multi-RREH CO$\mathcal{C}_r$4-valorization paper expresses the same idea as a linear program over GBOML nodes and hyperedges, minimizing the sum of annualized CAPEX and OPEX over two full years with hourly resolution (Dachet et al., 2023).

Capital annualization is standardized through a weighted average cost of capital. One formulation uses

$\mathcal{C}_r$5

with $\mathcal{C}_r$6 and technology-specific lifetime $\mathcal{C}_r$7 (Dachet et al., 2023). In the North Africa–Europe methane study, setting $\mathcal{C}_r$8 lowers the levelized methane cost from about $\mathcal{C}_r$9 to $\mathcal{E}_r$0, while installed capacities change very little; this makes financing cost a first-order driver for capital-intensive RREHs (Berger et al., 2021).

RREH planning is also shaped by regulation. The RFNBO study models additionality, temporal correlation, and geographical correlation through constraints that tie electrolyser consumption to local or PPA-backed renewable generation (Langer et al., 2023). Under those assumptions, the model always invests in the maximum electrolyzer capacity allowed when PPA investment is permitted, and the best business case is e-methanol with limited storage to secure hydrogen supply to the synthesizer (Langer et al., 2023). This suggests that, in regulated “green fuel” settings, the hub is not an unconstrained electricity arbitrage device; it is a tightly coupled RES-plus-P2X system.

For offshore and isolated RREHs, detailed storage modeling becomes central. The REMO model minimizes lifetime cost over pre-commissioning, capital, O&M, and decommissioning terms, and its REMO–DNN-BD extension adds a deep neural network-based battery degradation term,

$\mathcal{E}_r$1

(Toms et al., 8 May 2025). In the California coastal town case, the DNN-BD iteration yields a total cost that is about $\mathcal{E}_r$2 lower at iteration 3 than at iteration 1, illustrating that storage-aging economics can materially change the optimal RREH design (Toms et al., 8 May 2025).

A broader modeling framework for “Hubs for Circularity” argues that effective energy management requires combining objective functions, uncertainty representation, operational flexibility, and market participation with industrial-symbiosis-specific factors such as the type of symbiosis, the degree of information sharing, and collaboration structures (Nielsen et al., 27 Aug 2025). A plausible implication is that RREHs with extensive byproduct use, waste-heat cascades, or local co-use of water, oxygen, or synthetic fuels may require richer formulations than pure energy-cost minimization.

4. Monitoring, communication, and control

Operationally, RREHs require layered monitoring and control. At the lowest-cost end, a remote solar monitoring system was implemented with a Morningstar TriStar-45 charge controller, an Arduino/Freeduino-class microcontroller, a Motorola W260g feature phone used as a DTMF modem, and a planned server-side application (Wolfe, 2015). The field node polls the controller every 30 seconds for battery voltage, panel voltage, charge current, load current, and cumulative kWh; it validates data, maintains a small queue, answers incoming calls, outputs queued data frames as DTMF, and triggers repeated alarm calls when battery voltage falls below a preset cutoff (Wolfe, 2015). The paper explicitly states that this is not a complete RREH solution, but it offers “a concrete architecture pattern for ultra-low-cost monitoring” and “a hardware/software design style that prioritizes openness, universality, and field maintainability” (Wolfe, 2015).

For larger hubs, the communication stack becomes hierarchical. A wireless-assisted hierarchical framework proposes a regional aggregated controller co-located with or logically associated to a base station, direct links for stationary inverter-based resources (IBR–CUs), and device-to-device links for mobile or flexible resources (IBR–D2Ds) (Ge et al., 2022). Within each cell, the base station runs a three-step resource-allocation procedure: feasible reuse-partner detection, pairwise power allocation, and a Hungarian-algorithm-based matching stage (Ge et al., 2022). In the simulation setup, the carrier frequency is 2 GHz, the bandwidth is 4 MHz, the cell radius is 500 m, and the latency requirement is 20 ms (Ge et al., 2022). This communication layer is coupled to a two-layer control hierarchy: fast-timescale centralized control within each region and inter-regional distributed control across regional controllers (Ge et al., 2022).

At the network-control layer, distributed multi-horizon MPC provides a bridge between centralized optimality and decentralized privacy. The distributed model predictive controller for a network of energy hubs uses consensus ADMM to coordinate peer-to-peer electricity and heat trading while preserving local information (Behrunani et al., 2023). The multi-horizon extension increases the prediction horizon without keeping a fine discretization over the entire horizon, and in the benchmark three-hub network it shows superior performance in terms of total cost, computational time, and robustness to demand and prices variations (Behrunani et al., 2023). Experimental validation on a complete network with one real NEST energy hub and multiple virtual hubs confirms lower total network cost than islanded operation in two topologies: in Experiment 1, cost changes from $\mathcal{E}_r$3 CHF in islanded operation to $\mathcal{E}_r$4 CHF under MH-DMPC; in Experiment 2, cost changes from 645 CHF to 434 CHF (Behrunani et al., 2023).

Taken together, these studies indicate a continuum: from field-maintainable monitoring nodes, to hub-level regional controllers with adaptive wireless networking, to distributed MPC for coordinated hub networks. For RREHs, the appropriate point on this continuum depends on the number of assets, communication quality, ownership fragmentation, and control timescale.

5. Representative techno-economics and comparative performance

The most developed RREH literature concerns synthetic-fuel export. In the Algeria–Belgium comparison of four carriers, delivered costs at the Belgian gate are 150 $\mathcal{E}_r$5 for methane, 107 $\mathcal{E}_r$6 for ammonia, 120–121 $\mathcal{E}_r$7 for hydrogen, and 143 $\mathcal{E}_r$8 for methanol, all on an HHV basis (Larbanois et al., 10 Jul 2025). The same study reports overall efficiencies of 44% for methane, 60% for ammonia, 55% for hydrogen, and 43% for methanol, and concludes that ammonia demonstrates the most favourable cost-to-energy exported ratio (Larbanois et al., 10 Jul 2025).

Carrier	Delivered cost	Overall efficiency
CH$\mathcal{E}_r$9	150 $\mathcal{I}_r$0	44%
NH$\mathcal{I}_r$1	107 $\mathcal{I}_r$2	60%
H$\mathcal{I}_r$3	120–121 $\mathcal{I}_r$4	55%
CH$\mathcal{I}_r$5OH	143 $\mathcal{I}_r$6	43%

CO$\mathcal{I}_r$7 sourcing is a second major differentiator. In the Morocco–Belgium methane chain, the delivered cost is 157.93 $\mathcal{I}_r$8 for “DAC in Morocco,” 136.04 $\mathcal{I}_r$9 for “PCCC in Morocco,” and 145.51 $\mathcal{B}_r$0 for “PCCC in Belgium + DAC in Morocco” (Fonder et al., 2023). The corresponding global ESC efficiencies are 48.44%, 51.10%, and 51.26% (Fonder et al., 2023). The lowest-cost option is therefore the one using a nearby post-combustion source in Safi, but the paper also states that this route should be interpreted as a transition strategy because the coal plant is a finite-lifetime emitter (Fonder et al., 2023).

The multi-RREH CO$\mathcal{B}_r$1-valorization study reaches a similar conclusion at larger scale. In Scenarios 1–4, only the Algeria hub is deployed; Greenland remains unused because it is more expensive due to wind-only generation and higher flexibility needs (Dachet et al., 2023). The resulting CH$\mathcal{B}_r$2 costs are 136.00, 137.19, 133.89, and 129.27 $\mathcal{B}_r$3 in Scenarios 1–4, compared with 192.00 $\mathcal{B}_r$4 in Scenario 5 when only Greenland is available (Dachet et al., 2023). The shadow price of the zero-net-emission cap is about 162.77 $\mathcal{B}_r$5 in Scenarios 1 and 2 and about 235.65 $\mathcal{B}_r$6 in Scenario 5 (Dachet et al., 2023). This study also reports that “PCCC + CO$\mathcal{B}_r$7 transport is more cost-effective than DAC at hubs,” with an approximately 9.2% total system cost reduction versus DAC-only RREH designs (Dachet et al., 2023).

Remote-community and offshore interpretations show a different but related performance picture. In the REMO study for fully renewable offshore microgrids, 20-year lifetime costs for an offshore oil and gas platform range from 478.2 M$\mathcal{B}_r$8 in Texas, 728.9 M$\mathcal{B}_r$9 in New Jersey, with optimized BESS capacities from 66,300 to 171,600 kWh (Toms et al., 8 May 2025). For a large coastal town, lifetime costs range from 108.6 M$\mathcal{O}_r$0 in Florida, with BESS capacities from 31,200 to 66,300 kWh (Toms et al., 8 May 2025). The same study finds that wave energy converters become included in Alaska only when lifetime cost is reduced to 13% of current, while tidal energy converters become competitive in Florida only when cost is reduced to about 10% of current (Toms et al., 8 May 2025). This suggests that, under current techno-economics, OWT + FPV + BESS is the robust default architecture for offshore RREHs.

6. Limitations, misconceptions, and future directions

Several recurring limitations are explicit. Many optimization models assume perfect foresight, static capacities, a single central planner, and linearized technology representations (Berger et al., 2021, Dachet et al., 2023, Fonder et al., 2023). The taxonomy is “primarily qualitative and structural,” and does not yet formalize financing models, tariff structures, or profit-sharing (Dachet et al., 10 Jul 2025). The low-cost monitoring system has no completed server-side implementation and does not address encryption or authentication; the DTMF-over-GSM channel is inherently insecure (Wolfe, 2015). The wireless hierarchical framework focuses on communication and aggregation, and does not derive detailed inverter dynamics or stability under variable delays (Ge et al., 2022).

These limits matter because common misconceptions often treat RREHs as if they were only a siting problem. The literature points the other way. An RREH is not merely “renewables in a remote place”; it is a tailored graph of technologies, storages, imports, exports, byproducts, and local opportunities (Dachet et al., 10 Jul 2025). Nor is every offshore collector station an RREH in the stricter electrical sense: projects without both local multi-GW aggregation and multiple interconnections at the same node fall outside the Electrical Energy Hub definition (Bastianel et al., 8 Apr 2025). Likewise, low carrier cost at the destination does not eliminate the importance of process flexibility, storage degradation, control interoperability, or communication resilience.

Future work in the cited papers is correspondingly broad. Proposed directions include more RREHs and more energy-demand centres, other e-fuels such as ammonia and methanol, improved Greenland assumptions, coupling to detailed electricity markets and network models, full cyber-physical co-simulation, field validation at RREH scale, standardized data models and APIs, financial components in the taxonomy, and explicit social and environmental indicators (Dachet et al., 2023, Ge et al., 2022, Wolfe, 2015, Dachet et al., 10 Jul 2025). The H4C review adds a methodological warning: despite decentralized exchanges and multi-actor governance, the literature still shows a persistent reliance on centralised model structures (Nielsen et al., 27 Aug 2025). A plausible implication is that future RREH research will need to align physical architecture, communication topology, and optimization structure more tightly than current single-planner formulations do.