June 5, 2026 · 7 min read · Case Studies
A solar farm in the desert has no fiber. A wind farm 50km offshore has no cell tower. Both need to feed real-time telemetry into grid balancing systems that make decisions in sub-100ms windows. LTE-M covers the solar array. Private 5G covers the wind turbines. The SIM is the only component that crosses both domains.
A 100MW solar farm in the Arizona desert generates 2 million data points per day — panel temperature, irradiance, inverter status, power output per string. A 200MW offshore wind farm in the North Sea generates 5 million — blade pitch, gearbox vibration, generator temperature, wind speed at nacelle height. Both feed into a grid balancing system that must match supply to demand every 4 seconds. The solar farm has no fiber connection — the nearest POP is 30km away. The wind farm has no cell tower — it is 50km from shore. Both use cellular IoT. But the SIM architecture for each is completely different.
A utility-scale solar farm spans hundreds of hectares. Panel-level sensors — temperature, irradiance, voltage — are stationary, low-data-rate, and battery-powered. This sounds like NB-IoT territory. But solar farms have one requirement that eliminates NB-IoT: firmware updates. A solar inverter's control firmware is 2-5 MB. On NB-IoT at 20 kbps effective throughput, this takes 15-40 minutes — during which the inverter is offline. On LTE-M at 300-500 kbps, the same update takes 40-90 seconds. Across a 500-inverter farm, the difference is 50 hours of cumulative downtime vs 3 hours.
The SIM architecture: multi-IMSI with the two strongest carriers at the site location. Solar farms are built where land is cheap — which means rural, which means carrier coverage is unpredictable. A site survey with AT+CSQ at each planned inverter location determines which carriers have usable signal. The SIM carries profiles for the top two. The gateway aggregates panel-level data over LoRaWAN or wired Modbus, then backhauls over LTE-M to the SCADA cloud.
Offshore wind turbines are effectively industrial facilities on stilts in the ocean. Each turbine houses a nacelle with: generator temperature sensors (10+), gearbox vibration accelerometers (6-8), blade pitch controllers (3 per blade), yaw system monitors, tower structural sensors, and a meteorological station at hub height. Total: 50-80 sensor points per turbine. Data volume: 500 MB to 2 GB per turbine per day.
The connectivity challenge is not data rate — it is latency and reliability for grid balancing. When grid frequency deviates from 50 Hz by more than 0.2 Hz, the wind farm must respond within 4 seconds — adjusting blade pitch, reducing output, or disconnecting turbines. This requires URLLC-grade latency (sub-10ms) from the grid operator's control system to the turbine controller. Public cellular networks cannot guarantee this. Private 5G (n77/n78 band, 3.7-3.8 GHz) deployed on the wind farm's own infrastructure can — with dedicated spectrum, local breakout, and deterministic scheduling.
The SIM for private 5G is an eUICC with two profiles: a private 5G profile for grid-balancing traffic (sub-10ms, on-prem MEC) and an LTE-M public MNO profile for SCADA telemetry, firmware updates, and crew communications. Satellite (Iridium Certus or Starlink Maritime) serves as the backhaul failover when the subsea fiber to shore is damaged.
Trenching fiber to a solar farm costs $15-50 per meter depending on terrain. For a solar farm 30km from the nearest point of presence, that is $450,000 to $1.5 million. LTE-M connectivity for 500 inverters at $1-3/month per SIM = $6,000-18,000/year. The fiber installation cost alone equals 25-80 years of cellular connectivity. For wind farms, submarine fiber to shore costs $50-200 per meter — for a farm 50km offshore, that is $2.5-10 million. Private 5G with satellite backup is $50,000-200,000 in infrastructure plus $500-2,000/month in satellite airtime.
The procurement crossover: if the renewable asset's operational life is 20-30 years, fiber may amortize. If the asset is a 5-10 year project with potential relocation (common for solar), cellular wins decisively. The SIM architecture should be project-scoped — not catalogue — because each site's carrier coverage, latency requirements, and physical accessibility are unique.
Source: Kigen, "5 ways eSIMs accelerate grid modernization for C&I utilities", November 2025. Available at https://kigen.com/resources/blog/5-ways-esims-accelerate-grid-modernization-for-ci-utilities/
Source: PUSR IoT, "How IoT Routers Solve Remote Equipment Networking Dilemmas", February 2026. Available at https://www.pusr.com/blog/How-IoT-Routers-Solve-Remote-Equipment-Networking-Dilemmas