Our team at General Power works with data center backup power systems every day. Every one of our technicians and engineers is certified by the Electrical Generating Systems Association (EGSA), so our training is backed by hands-on field experience. When clients ask how data center generators work, we answer from what we have seen during real outages.
A data center generator is a standby backup power system. It starts on its own when the utility grid goes down and can run for hours or even days until grid power comes back. Below, we walk through the outage sequence, redundancy, and power distribution the same way our engineers explain them to clients.
Key Takeaways
- A data center generator is a standby system that starts automatically when the grid fails and can run for hours or days.
- The UPS powers servers instantly during an outage, giving the generator time to start.
- The automatic transfer switch moves the load to the generator only after its voltage and frequency are stable.
- N+1 and 2N redundancy designs keep a single equipment failure from causing downtime.
- Running generators above about 30% load prevents wet stacking and carbon buildup.
- General Power supplies data centers with diesel generators from 10 to 2,000 kW and transfer switches from 100 to 3,000 amps.
The Core Components of a Data Center Backup Power System
A generator does not work alone. It is one part of a larger system, and each part has a specific job. If one piece fails or is set up wrong, the whole system can let the facility down when it matters most.
We focus on four main parts: the generator, the uninterruptible power supply, the automatic transfer switch, and the power distribution equipment. Here is what each one does and how they work together.
The Generator Engine and Alternator
Most of the systems we work with run on heavy-duty diesel, but we also see natural gas units and turbine-based systems. Each fuel type has its own strengths, and our team helps clients compare diesel and gas options for their site.
|
Fuel Type |
Where We See It Most |
Main Strength |
Main Trade-Off |
|
Heavy-duty diesel |
Most data centers |
Fast startup and fuel stored on site |
Fuel must be tested and maintained |
|
Natural gas |
Sites with a strong gas supply |
Clean burning, no large tanks |
Depends on the gas pipeline |
|
Turbine-based |
Large facilities |
High power from a compact unit |
Higher cost and slower startup |
The engine burns fuel, which turns the fuel’s chemical energy into mechanical rotation. That rotation spins the alternator, where a magnetic field passes through copper windings to create electricity the facility can use.
The Uninterruptible Power Supply (UPS)
The UPS bridges the gap between the moment the power goes out and the moment the generator takes over. A generator needs time to start, but servers cannot go without power for even a split second. The UPS steps in instantly so the equipment never notices a problem.
A UPS stores energy in batteries or flywheels, which are heavy spinning wheels that hold energy in their motion. Most data centers use a double-conversion UPS. It turns incoming AC power into DC and back into clean AC all the time, so servers get steady, conditioned power with no delay.
The Automatic Transfer Switch (ATS)
The ATS is the decision maker of the system. It works with monitoring equipment and generator control panels that watch utility power around the clock, and when something goes wrong, it signals the generator to start.
Once the generator is ready, the ATS moves the load from the utility to the generator, then back again when the grid returns. A sticking switch can leave a facility running on batteries alone, so we check the ATS closely on every service visit and help clients pick the right transfer switch for their load.
Power Distribution Equipment
Switchgear is the main control point. It routes power, protects circuits with heavy-duty breakers, and lets technicians safely isolate parts of the system for repair.
Transformers change the voltage to levels the equipment can use. Power Distribution Units (PDUs) and Remote Power Panels (RPPs) then split the power into smaller circuits that feed each server rack.
How Data Center Generators Work During a Power Outage: Step by Step
When the grid fails, the backup system follows a set sequence that usually takes less than a minute from outage to full generator power. Each step has to happen in the right order for the servers to stay online.
Our engineers compare it to a relay race. The utility hands off to the UPS, the UPS hands off to the generator, and the generator hands back to the utility once the grid is stable.
Step 1: Detecting the Outage
The monitoring equipment, the ATS, and the generator’s auto start control panel are always watching the incoming power. They start the backup process when they see any of these conditions:
- The utility power drops.
- The voltage fluctuates too high or too low.
- The grid fails completely.
The ATS waits a brief moment to confirm the problem is real and not just a quick flicker. If the problem continues, it sends a start signal to the generator, which keeps the unit from starting every time the power blinks.
Step 2: The UPS Bridges the Gap
Because a double-conversion UPS is always feeding the servers, there is no gap when the grid fails. The IT equipment never loses power, not even for a moment.
A typical UPS can carry the load for a few minutes up to about 15 minutes. That is plenty of time for a healthy generator to start, which is why we stress regular battery testing. Weak batteries shrink that safety window.
Step 3: Generator Engine Startup
The engine cranks and speeds up to its operating speed, often around 1,800 RPM for generators in the U.S. Engines kept warm with block heaters start faster and more reliably.
Startup typically takes about 10 to 60 seconds. During this time, the energy chain kicks in: fuel burns to create mechanical rotation, and the AC alternator turns that rotation into electricity.
Step 4: Transferring the Load to Generator Power
The ATS does not switch the load the instant the engine starts. It waits until the generator’s voltage and frequency are stable, because moving the load too early could damage equipment or stall the engine.
Then the critical loads move over: IT server racks, cooling systems, and other support systems like lighting and security. Cooling is a big load, and keeping it running matters just as much as keeping the servers on.
Step 5: Returning to Utility Power
When the grid comes back, the system does not rush to switch over. It first confirms that utility power stays steady for a set amount of time, which guards against the grid failing again right away. Once the grid passes these checks, the shutdown routine follows:
- The ATS confirms grid power is back and stable.
- The load switches back to the utility.
- The generator runs a no-load cool-down, then shuts off.
- The UPS batteries recharge for the next event.
The cool-down step lets the engine’s temperature drop slowly, which protects its internal parts from heat damage. Once the batteries are full again, the whole system is back on standby and ready for the next outage.
Redundancy: How Data Centers Prevent Generator Failure From Causing Downtime
Any machine can fail, and a generator might fail right when it is needed. That is why data centers plan for failure from the start. Redundancy means having extra equipment ready to take over.
The two setups we see most are N+1 and 2N. In both cases, “N” stands for the amount of equipment needed to carry the full load.
N+1 Configuration
N+1 means one more generator than the load requires. If a site needs four 650 to 2,000 kW generators to run, it has five, so one unit can fail without causing an outage.
It also lets our technicians take one generator offline for routine maintenance without leaving the facility unprotected. For many clients, N+1 offers a strong balance of protection and cost.
2N Configuration
2N means two full, independent power paths. Each path has its own generators, UPS units, and distribution gear, and either one can carry the whole load by itself.
This complete duplication gives maximum resilience. It is common where even a short outage would be very costly, such as banks, hospitals, and large cloud providers.
Choosing Between N+1 and 2N
Our engineers help clients weigh cost, reliability, and maintenance flexibility. The table below shows how the two designs compare.
|
Factor |
N+1 |
2N |
|
Cost |
Lower, one extra unit |
Higher, doubles the equipment |
|
Reliability |
Covers one generator failure |
Covers a full power path failure |
|
Maintenance flexibility |
Service one unit at a time |
Service an entire path at once |
|
Best fit |
Most enterprise data centers |
Mission-critical facilities |
There is no single right answer for every site. We look at each client’s uptime goals, budget, and growth plans, then size each generator correctly before recommending a design.
Load Management and Generator Health
Many people think a generator that runs lightly is a happy generator. In our experience, the opposite is often true, because diesel engines are built to work hard.
This matters most during routine generator exercise and long outages when the IT load is low. Good load management keeps the generator healthy and ready for the next emergency.
Why Minimum Load Matters
Our engineers aim to keep generators above roughly 30% of their rated load while running. At that level, the engine gets hot enough to burn its fuel completely and stay clean inside.
When a generator runs too lightly, the engine stays cool and fuel does not fully burn. To prevent this, we often use load banks during testing, which add an artificial load so the generator can run at a healthy level.
Common Engine Problems From Underloading
Wet stacking happens when unburned fuel and oil pass into the exhaust system. Signs include black, oily drips from the exhaust, heavy smoke, and lower power output.
Carbon buildup is the second problem. Hard deposits form on injectors, valves, and piston rings over time, leading to poor performance, higher fuel use, and costly repairs that can shorten engine life if they are not caught early.
How Generator Power Reaches the Servers
Making electricity is only half the job. That power must reach each server at the right voltage, or equipment can be damaged or shut down.
Many of our service calls start here. A healthy generator does no good if the distribution path has a weak link, and problems like a failing voltage regulator can send unstable power downstream.
Stepping Down Voltage
Power enters a facility at a higher voltage because it travels more efficiently with less energy lost as heat. That voltage must be reduced before it reaches sensitive equipment.
The path runs from switchgear to step-down and isolation transformers, then to the UPS and the PDUs. Industry guidance from the Uptime Institute and equipment makers such as Schneider Electric and Eaton supports this layered design, which lowers voltage and adds protection at each stage.
Common Data Center Voltage Levels
Most U.S. data centers use 480V three-phase power for facility-level distribution to feed UPS systems, cooling units, and main panels. Server equipment usually runs on 208V or 120V.
The step-down to 208V or 120V typically happens in transformers inside UPS units or in PDU and RPP gear. Placing transformers close to the racks saves energy and keeps the system easier to manage.
How General Power Supplies Data Centers With Diesel Generators
General Power Limited, Inc. is a Miami, Florida-based Original Equipment Manufacturer (OEM) and master distributor of power generation products. We give data centers a single source for diesel generators, automatic transfer switches, alternators, and parts. Here is what we bring to every data center project:
- 100+ diesel generators (10 to 2,000 kW), including EPA-compliant models, ready to ship.
- 100+ automatic transfer switches (100 to 3,000 amp) to manage the handoff between utility and generator power.
- Same-day quoting and shipping, plus customization of generator packages, accessories, and upgrades.
- Equipment purchased directly from original manufacturers, with products individually serialized and traceable to the factory.
With an A+ rating from the Better Business Bureau and customers in 100+ countries on every continent, we have the stock to move fast when a data center project cannot wait. Our EGSA-certified sales specialists and engineers make ordering your next generator simple and match each unit to your facility’s load and redundancy plan.
Understanding How Data Center Generators Work
A data center’s backup power is built in layers, and each one has a clear job. The UPS covers the first seconds, the generator provides long-term power, and the ATS manages the handoff between them. Redundancy guards against equipment failure, and the distribution system delivers the right voltage to every rack.
When all of these layers work together, they deliver one goal: zero interruption to critical IT loads. Contact our EGSA-certified team today to talk about the backup power your data center depends on, from new generators to ongoing maintenance and testing.