
Telecom resilience is usually discussed as an engineering problem. In Ukraine it is a capital-allocation problem with a measurable failure point — and the operators have published the number.
During blackouts in Chernihiv and Sumy, Vodafone Ukraine reported average network availability of 97% and 99% respectively. Those are the figures that get quoted.
The more useful figure is the other one in the same disclosure: in Chernihiv, after eight hours running on batteries, availability fell to 75%.
That is the honest shape of the problem. Backup power does not make a network outage-proof. It buys a defined number of hours, and the cost of each additional hour rises steeply.
Vodafone Ukraine — wholly owned by NEQSOL Holding — put UAH 754 million into network energy resilience in 2024 alone. Over a two-year period the company has said it spent roughly UAH 2 billion on rechargeable batteries. Network investment rose about 40% year on year.
By the end of 2024 every base station had backup power. Around 65% could run between four and 72 hours unaided.
The engineering choices are specific. The operator moved to high-capacity lithium-ion batteries that recharge two and a half to three times faster than the previous generation — which matters when the grid returns for short, unpredictable windows. It also increased the number of base stations backed by diesel generators nearly fourfold.
In a normal grid failure, you size batteries for the expected outage. In Ukraine the pattern is different: rolling, repeated outages with brief restoration windows in between.
Under that pattern, the binding constraint is not how long a battery lasts. It is whether it can refill before the next cut. A battery that runs eight hours but needs twelve to recharge is a battery that is empty by the third day of a schedule.
That is an unusual spec to optimise for, and it is the kind of operating knowledge that only comes from running infrastructure under sustained attack.
Russian strikes on the power system between October 2025 and January 2026 produced a peak-hour capacity deficit of 5–6 GW. That is the environment the network has to work inside — not an occasional fault, but a structural shortfall.
The response has moved from backup to generation. Vodafone has now completed solar installations at 100 base stations across 14 regions, expected to supply roughly 30% of those sites’ energy from renewables. Sites run on solar during daylight and store the surplus in lithium-ion batteries.
Hundred-site solar is not a decarbonisation programme at this scale. It is an autonomy programme that happens to be renewable.
Resilience is not free, and in a market with falling real incomes the cost has to land somewhere. Both major operators are reported to be raising tariffs ahead of the winter, citing the energy crisis and higher operating costs — with Vodafone increases reported to take effect on 8 October 2026. That date appears in a single report and CBJ has not independently confirmed it.
This is the uncomfortable part of the resilience story. The investment is real and the engineering is sound, but the funding model is partly the subscriber base of a country at war.
Three things follow.
Capex is no longer discretionary. Energy resilience spending competes with spectrum, coverage and 5G readiness for the same capital. A network spending UAH 2 billion on batteries is a network deferring something else.
Operational know-how has become an exportable asset. Very few operators anywhere have run a national network through four years of sustained infrastructure attack. The knowledge of how to do it — battery chemistry under rolling outages, generator logistics, site autonomy design — has value well beyond Ukraine.
It sits inside a group making long-horizon infrastructure bets. The same owner is building the Kardesa submarine cable across the Black Sea with Vodafone Group, and holds the AzerTelecom side of the Trans-Caspian link. Wartime network operations and submarine cable construction draw on the same balance sheet.
The eight-hour wall. Current architecture degrades materially past roughly eight hours of continuous outage. A harder winter than 2025–26 would test that directly.
Tariff increases have a ceiling. Raising prices into a wartime economy works until it does not. Churn and non-payment are the limits.
Physical risk is uninsurable at normal rates. Base stations, fibre and power infrastructure remain targets. No amount of battery capacity addresses a destroyed site.
Three measurable things this winter: whether availability holds above 90% during extended outages, whether solar expands past the first 100 sites, and whether the tariff increases hold without visible subscriber loss. Those will show whether resilience at this cost is sustainable or simply survivable.
Sources: Vodafone Ukraine disclosures on network availability in Chernihiv and Sumy; Interfax-Ukraine; Open4Business; dev.ua; reporting on winter 2026 tariff adjustments. Figures as published; the 8 October 2026 tariff date is single-sourced and unconfirmed.