Designed UPS availability*
Demonstrated module swap*
True online VFI efficiency*
These are the requirements the estate is already held to, applied to the AC power chain rather than around it.
Traffic grows and the site outgrows its plant. A declared ceiling, reached in steps under load, makes the next increase a change of size, not of system.
A module fails at three in the morning. The reserve holds the site while the repair happens, so how fast the repair is decides how much reserve you need.
The grid sags and recovers. Output held inside tolerance through the dip, and a controlled current draw on the way back, keeps the load unaware it happened.
The generator run is scheduled and the load stays on. Testing at full load with the site live turns backup from something you believe into something on record.
Power is behind about one network incident in seven, and more than half of all lost service hours.*
It rarely has its own line on the risk register. DC and AC plant keep separate records.
Both can change. So can the shared part that takes redundancy down with it.
DARA™ distributes what centralised designs concentrate. Every module carries its own rectifier, inverter, static bypass and control, and decides with the others rather than waiting on one board.
No single control board. No single static bypass. No shared parallel bus. Each module is a complete uninterruptible power supply, so a fault has nowhere to travel beyond the module holding it.
DDM™ means no component decides for the system. Modules assess conditions, exchange state and act together. Lose one and the rest redistribute its capacity, instead of the whole system stopping because a central control did.
One of the numbers this estate is specified against. A module is isolated, swapped and tested inside the running frame before it takes load. Demonstrated in under 3 minutes, no transfer to static bypass.*
More than 100 measurement points report module, battery and system status over standard SNMP, the same protocol the operator’s monitoring already reads. Connecting it into the NOC is the integrator’s work.
Double conversion handles the sag, and the load never sees it. The harder moment is restoration, when a site’s rectifiers, chargers and cooling all reach for current at once.
Across a distributed estate the battery is what gets replaced, and the replacement is a site visit before it is a purchase. Cycle life, footprint and how often a technician has to attend are the numbers that compound.
StratusPower™ and CumulusPower™ cover three-phase from 10 kW to 3.75 MW, with DARA™ throughout. StratusPower™ is also available as a Universal Rack UPS in 10 to 30 kW modules, fitting any 19-inch rack up to 120 kW.
Up to 1 MW/m² in a transformerless frame, with modules light enough for two people to handle. Where there is no room for another cabinet and the floor will not take more weight, density is what makes the project possible.*
Architecture turns into operations. These are the changes that show up after the site is running.
A module swap replaces a system intervention, so a technician travels on a schedule rather than on an alarm. Across an estate, visits are what compound.
Battery status is visible per site, so reserve time is a number you compare across the estate instead of one that only becomes real during a discharge.
Modules match the load rather than the nameplate, so the energy line follows what the site draws. On sites running below capacity, that is where it shows.
One system across sites means one spares holding, one training set, one set of numbers. A framework can be written once and called off per site.
Interface behaviour, figures at defined load points, and an architecture a specification can carry. If it is not specified, it is not evaluated.
What it does to the energy line, to maintenance, and to the sites you would otherwise have to re-engineer. Arguments that hold in a capital review.
Modernising power under a network that does not stop. Staged capacity, module-level work, and no service window to negotiate with operations.
Architecture and operation. Both documented.
The engineering that removes failure paths before the site carries traffic.
Designed for up to 99.9999999% availability, modelled on the topology described above.*
EN/IEC 62040-1, -2 and -3. Input voltage window declared according to VFI SS 111, voltage and frequency independent.
Hot-swap under live load with no bypass transfer. The replacement time is a demonstrated procedure, not a specification figure.*
Input power factor 0.99, output unity. THDi at or below 0.9% linear. MTBF on request. Autonomy sized to the site.
Centiel was created to solve a structural problem in mission-critical power: systems that only reveal their weaknesses once they are live. This is the engineering intent behind every architecture decision that follows.
The operators are not named. The agreements that made the work possible are why.
Megawatt-scale installations inside national telecom operators across Europe, the Middle East and Asia.*
Telecom projects are delivered through partners who carry installation, commissioning and ongoing maintenance.
Not a startup, not a rebrand. Swiss-made, on an engineering line running back 45 years to our co-founder’s transformerless UPS design.
Financial transparency and institutional accountability. The governance structure of a company built for 30-year infrastructure partnerships.
The Telecom Power Specification Guide takes the AC side of a network site from requirement to tender: clause wording for reserve, repair and architecture, the parameters that make offers comparable, and the sequence for modernising a live site.
The guide covers the AC uninterruptible power supply beside the DC plant, not the plant itself, and names each standard for what it governs. One document, no sales call attached.
Wording for reserve, time to repair a power-source failure, and what the modules are allowed to share. The numbers a site is selected on.
Availability is MTBF over MTBF plus MTTR. Time to repair is the second term, which is why it belongs in the specification, not after it.
Power factor, THDi, efficiency at defined load points, MTBF and autonomy, with a column for what each vendor publishes. Ours included.
What happens to the load at each step, and what has to be true before the first module goes in. The full procedure travels with the order.
Performance figures are based on architectural modelling and configuration-specific assumptions. Actual performance depends on system configuration, installation environment, maintenance practices, and operating conditions.
Module replacement. Demonstrated live module replacement in under 3 minutes, with the system under load, no transfer to static bypass, and the module isolated and tested within the running frame before accepting load. A standard maintenance procedure, subject to trained personnel, site procedures and spare-module availability.
Efficiency. Up to 97.6%, depending on configuration and operating conditions.
Availability. Availability design target of up to 99.9999999%, based on architectural modelling and configuration-specific assumptions.
Service response. Response time available under applicable service agreements and defined regional coverage.
Lead time. Typical factory lead time is four to six weeks from confirmed order and final technical release, subject to configuration, production capacity and prevailing supply conditions.
Telecom deployments. Centiel modular UPS systems are in production at multi-megawatt scale with telecom operators across Europe, the Middle East and Asia. Operator names are held in confidence. Deployment detail is shared on request.
Incident proportions. Figures describe reported telecommunications incidents in European regulatory reporting and are not a statement about any individual operator or market.
Scope. Capacity, power density, efficiency and availability figures apply to Centiel’s modular UPS ranges.