Skip to content
INTERNET INFRASTRUCTURE & NETWORKS

Powering Remote Internet Infrastructure with Solar Energy: A Practical Blueprint for Network Operators

As the global Internet expands its reach into increasingly remote and challenging locations, it simultaneously carries a vastly growing share of everyday mission-critical services. From essential telephony and messaging networks to modern banking systems and vital telemedicine applications, maintaining an uninterrupted and reliable power supply has transformed from a secondary operational consideration into an absolute necessity for both end users and network infrastructure providers alike.

In response to these expanding demands, solar photovoltaics combined with advanced battery energy storage systems have emerged as a highly attractive alternative to traditional long and precarious electrical grid supply lines, as well as costly, high-maintenance diesel generators. Over the past decade particularly, the economics of renewable energy have shifted dramatically. Solar panel manufacturing prices have plummeted to a fraction of their historical costs. Concurrently, highly reliable and high-capacity lithium iron phosphate batteries have become widely available in quantity at significantly lower price points. As a result, solar photovoltaic technology is now widely recognized as a worthwhile addition to both residential and commercial grid supplies, proving its value even within densely populated urban environments where reliability remains a concern.

Despite these advancements, navigating the adoption of solar photovoltaic technology can still prove bewilderingly complex for newcomers, network technicians, and local infrastructure planners. This inherent complexity frequently leads to poorly planned projects that ultimately fail to meet community needs, stall midway through implementation, or are abandoned prematurely. In worst-case scenarios, installed infrastructure is eventually removed entirely and never replaced. Dr. Ulrich Speidel, a senior lecturer in Computer Science at the University of Auckland whose connectivity research and implementation work has been supported by the APNIC Foundation, notes that he has witnessed such failures occur even within projects managed by established power utilities in developed urban areas. Therefore, remote operators who feel anxious or overwhelmed about making the transition are entirely justified in their caution. However, experts emphasize that renewable energy systems can be made to function reliably, provided stakeholders abandon unrealistic expectations and follow a disciplined approach to planning and execution.

Step 1: Figuring out what you need

Before reviewing pricing catalogs, requesting commercial quotes, or purchasing equipment, the fundamental preliminary task for any organization is to meticulously determine the exact equipment loads that the system will need to support. Network operators must account for a wide variety of hardware, including servers, routing equipment, network switches, uninterruptible power supply units, climate control systems, facility lighting, on-site computing terminals, link transmitters and receivers, and power tools utilized by maintenance working parties. Compiling this comprehensive inventory into a structured spreadsheet is widely considered an indispensable first step.

Every piece of electrical equipment features an associated power consumption rating, typically located on the manufacturer’s identification label, designated by a numerical value ending in the letter ‘W’ to represent watts. When physical labels are obscured or missing, operators can consult device documentation online, leverage artificial intelligence estimation tools, or utilize inexpensive plug-in power meters to measure actual draw. Crucially, planners must also remain vigilant regarding hidden loads, such as hot water cylinders or other heat-generating appliances that might be concealed behind utility cupboards or integrated into existing circuits. If these thermal loads cannot be physically isolated or excluded from the photovoltaic power supply, their energy requirements must be incorporated into the total calculations.

Once these individual power figures are compiled and aggregated within the planning spreadsheet, the resulting sum represents the absolute minimum power output required from the system during peak operational periods. For appliances that do not operate continuously—such as electric kettles or intermittent computing tasks—operators must estimate realistic average consumption or measure actual usage patterns over time. Completing this comprehensive power audit prevents organizations from procuring undersized systems incapable of supporting essential operations.

Following the power audit, operators must determine the specific system architecture best suited to their operational environment. The first primary option is a standalone, island photovoltaic system, which operates entirely independently without any connection to a broader electrical grid. This configuration is frequently the only viable choice for remote locations completely devoid of grid infrastructure. While standalone systems offer the least administrative and technical complexity—eliminating the need to negotiate grid-interconnection permissions with local utilities—they demand rigorous capacity planning. Operators must scale the system to sustain critical loads throughout extended periods of inclement weather or low solar irradiance. Consequently, this requires an oversized system that generates substantially more power than daily consumption requirements. Furthermore, because the system cannot export surplus energy into a local grid, there is no direct mechanism to accelerate financial return on investment. Any necessary maintenance requiring a full system shutdown also results in an immediate loss of power to connected loads, unless bridged by a mobile backup generator.

Solar PV for Internet resilience (Part 1): Designing your system | APNIC Blog

Alternatively, grid-connected photovoltaic systems offer the capability to temporarily island themselves during sudden utility outages, presenting a distinct set of operational advantages and disadvantages. These systems allow operators to export excess generation back into the local grid, offsetting initial capital costs. In smaller microgrids, such as those found on Pacific islands relying primarily on imported diesel fuel, exporting surplus solar power is particularly advantageous as it actively reduces operational load, mechanical wear on generators, and overall diesel consumption, while simultaneously contributing to local grid stability.

However, grid-connected setups introduce greater technical complexity, requiring management across three distinct power sources—photovoltaic panels, storage batteries, and the utility grid—and three potential destinations, including facility loads, batteries, and the grid. Determining the optimal distribution of energy between charging storage or exporting to the grid requires sophisticated modern inverter management systems. Additionally, grid-connected facilities mandate the use of an uninterruptible power supply for sensitive equipment that cannot tolerate momentary power interruptions. Because grid-tied inverters must strictly synchronize with the alternating current frequency of the utility grid, an unexpected grid failure forces the system to transition to island mode and generate its own independent frequency. This frequency will rarely match the returning grid signal when utility power is restored, necessitating a brief inverter shutdown phase at both the onset and conclusion of an outage, which must be bridged by backup hardware. Experts strongly advise against selecting photovoltaic installations solely for emergency backup purposes, noting that maintaining an idle solar system while continuing to purchase expensive standard grid power represents the least financially efficient operational strategy.

Step 2: Modelling your system

Following the completion of the power audit and architectural selection, the subsequent phase involves modelling the installation to determine the precise panel generating capacity required and the physical feasibility of accommodating that equipment on-site. Standard photovoltaic panels typically feature a peak generating capacity ranging between 430 and 460 watt-peaks. However, panels achieve these maximum ratings exclusively under optimal laboratory conditions, characterized by direct perpendicular sunlight on clear days, clean glass surfaces, and relatively cool ambient temperatures. Real-world environmental factors invariably reduce actual energy yields, underscoring the necessity of computer-aided system modelling.

To execute accurate modelling, planners utilize specialized software tools provided by reputable inverter manufacturers and industry organizations. These digital tools require precise inputs, including geographical location, roof pitch and orientation, available surface area, and potential shading obstructions from nearby vegetation or structures. For smaller installations supporting loads under 10 kilowatts, planners generally select a hybrid inverter capable of handling at least ten times the average power requirement calculated during the initial audit, ensuring sufficient capacity to simultaneously recharge storage batteries and supply active loads during limited daily sunlight windows. Larger installations require dedicated, segregated inverter capacities for both the solar array and the battery banks.

Regarding energy storage, experts recommend selecting a battery capacity sufficient to sustain baseline loads for up to five days without any active solar recharging. Storage capacity is quantified in kilowatt-hours, and dividing total available storage by anticipated load yields the expected autonomous backup duration. As a fundamental design principle, installations featuring larger panel arrays and higher inverter capacities can tolerate smaller battery storage volumes, as rapid generation enables faster recharging cycles.

During the physical modelling process, panels are connected in series to form electrical strings, elevating nominal direct current voltages to a target threshold approximately 100 to 200 volts below the maximum permissible limit for the specific installation site. These strings are then mapped to the maximum power point tracking inputs of the virtual inverter to evaluate performance under simulated island-mode conditions. By running annual simulations tailored to specific geographic coordinates, operators can analyze projected energy yields, identify generation surpluses or deficits, and iteratively adjust panel counts, inverter capacities, or battery volumes. This modelling phase allows organizations to identify non-essential loads that can be deliberately shed during prolonged outages, ensuring high confidence in system reliability before physical procurement commences.

Looking ahead to the subsequent phases of deployment, network operators who complete this rigorous preliminary assessment gain a comprehensive understanding of their true power consumption profile and a validated model outlining the necessary hardware capacities. Future phases of the initiative will address the critical processes of selecting reliable suppliers and equipment components, evaluating competitive commercial quotes, assessing cybersecurity protocols and technical support agreements, and implementing long-term maintenance strategies to guarantee sustained operational integrity for remote solar-powered infrastructure.