How UPS system available fault current affects disconnection times and how to get it right

A protective device can only disconnect quickly if enough fault current flows to operate it. That becomes a problem when a downstream fault is supplied by an inverter rather than the mains. Many UPS systems can deliver excellent continuity, but only a limited short-circuit current, often for a very short time. If the protective device has been selected as though the fault current will always come from the main low-voltage supply, the installation can fail to meet the required disconnection time when running on battery. We see this all the time on smaller UPS systems that have not been designed correctly.

This is not a minor design detail. It affects shock protection, fire risk, selectivity, test results, certification, and the credibility of the installation. It also catches people out because the same circuit may look compliant when the mains is present, then behave very differently when the UPS is supporting the load.

The available fault current changes when the inverter becomes the source.

Why low fault current is built into many UPS systems

A transformer-fed LV supply can usually provide high prospective fault current because the source impedance is low. Under fault conditions, current rises rapidly until the upstream protective device opens or the source voltage collapses.

A UPS inverter is different. It is a self-contained electronic system; not a large transformer supplied from the national grid. The inverter stage has semiconductor switches, DC link components, filters, control software, and thermal limits. These parts must be protected from destructive current.

For that reason, most UPS inverters are current limited. During a downstream short circuit, the inverter may deliver a multiple of rated current, but only within strict limits. The exact figure depends on the model, topology, rating, operating mode, and manufacturer. Some units may provide a high current for a few milliseconds to help clear downstream devices. Others may fold back, shut down the output, or transfer to bypass if conditions allow.

That behaviour is intentional. It protects the UPS. The problem is that downstream protective devices may not see enough current to operate within the required time.

UPS operating mode matters:

• Normal mode with mains healthy

  The load may be supplied through the inverter, with the rectifier fed from the mains. A static bypass may also be available. Depending on design, a fault may be supported by the inverter, by the bypass path, or by a transfer to bypass.

  
  

• Battery mode

  The inverter is the source. The mains is absent or outside tolerance. Fault current is usually limited to what the inverter and battery system can supply.

  
  

• Bypass mode

  The load is supplied from the bypass input. Available fault current may be similar to the upstream LV supply, subject to impedance and upstream protection.

  
  

• Maintenance bypass

  The UPS may be electrically bypassed. Fault levels can change again, often increasing because the inverter is no longer in the fault path.

  
  

The key point is simple: there is not one fault level for the installation. There are several, and the lowest one often applies at the worst time, during backup operation.

How low fault current affects disconnection times

Automatic disconnection of supply relies on the protective device operating quickly enough under fault conditions. In a conventional TN installation, a line-to-earth fault current must be high enough to operate the fuse, MCB, MCCB, RCBO, or other protective device within the maximum permitted time.

BS 7671 sets requirements for protection against electric shock using automatic disconnection of supply. For example, final circuits and distribution circuits have maximum disconnection times depending on earthing arrangement, nominal voltage, and circuit type. The exact design must be checked against the current edition of the Wiring Regulations, but the principle is stable: the earth fault loop impedance and protective device characteristics must work together.

The issue with a current-limited source is that the protective device may sit in the wrong part of its time-current curve.

Take a common example. A type C MCB needs a higher current to operate its instantaneous magnetic trip than a type B device of the same rating. If the UPS inverter can only deliver a limited multiple of rated output current, a downstream type C breaker may not trip magnetically. It may eventually trip thermally, but thermal operation can take seconds, tens of seconds, or longer depending on the current.

That can create three practical problems.

The circuit may not meet the required disconnection time

If a line-to-earth fault occurs downstream while the system is on inverter, the protective device must still clear the fault fast enough. If the available earth fault current is too low, the device will not disconnect within the required time.

This is where a design can pass a normal loop impedance test with mains present, then fail by calculation when assessed in UPS battery mode. The test has measured the upstream mains fault path, not the inverter-limited condition.

Selectivity can fail in unexpected ways

Low fault current can stop the nearest downstream device operating. Instead, the UPS may current-limit, shut down, or trip its own output protection. That can drop a whole board rather than one final circuit.

In critical supplies, this matters. A fault on one non-essential socket outlet should not take out controls, network equipment, emergency systems, or monitoring gear fed from the same backed-up distribution.

Faults can remain energised for longer

If the inverter continues supplying a limited current into a fault, exposed-conductive-parts can remain energised until something trips or the UPS shuts down. The current may be too low to operate an overcurrent device quickly, but high enough to create danger.

That is the uncomfortable part. Low available fault current sounds safer because the current is lower. For ADS, it can be worse, because the protective device needs current to disconnect.

MCB trip curves must be checked against inverter fault current, not just mains fault current.

Why normal fault calculations can be misleading

Many designs start with prospective fault current and earth fault loop impedance at the origin, then calculate downstream values using cable impedance. That works when the source behaves like a conventional supply.

A UPS introduces extra variables:

• inverter current limit

• duration of overload or short-circuit support

• bypass availability

• transfer logic

• battery state and DC link behaviour

• output transformer or transformerless design

• neutral and earth reference

• upstream and downstream protective coordination

• manufacturer protection settings

  
  

The quoted short-circuit capability of a unit must be read carefully. A data sheet might state a current for a short duration, a different overload rating for several seconds, and another behaviour under sustained fault. Those values are not interchangeable.

For example, a UPS might support a short burst of fault current long enough to clear a fast electronic protective device, but not long enough to operate a downstream breaker whose magnetic trip needs a higher current. Another unit may transfer to static bypass on overload, but only if the bypass supply is healthy and synchronised. If the mains has failed, that option disappears.

A design based only on the bypass fault level is incomplete. During battery operation, the bypass may not exist as a usable source.

There is also a difference between line-to-neutral short-circuit current and line-to-earth fault current. For disconnection under ADS, the earth fault path is critical. The earthing arrangement on the UPS output must be clear. A transformerless unit, a separately derived output, or an output isolation transformer can each change how earth faults behave.

Designers should confirm:

• whether the UPS output neutral is solidly referenced to earth in each mode

• whether the neutral-earth link changes between normal, battery, bypass, and maintenance bypass

• whether the output is TN-S, TN-C-S derived, TT, IT, or another arrangement

• how earth fault detection works inside the unit

• whether RCDs or insulation monitoring devices are required or prohibited by the manufacturer

  
  

This is not guesswork territory. The manufacturer’s installation manual and short-circuit data should be treated as design inputs, not afterthoughts.

A simple example of the problem

Consider a small backed-up distribution board feeding final circuits protected by 16 A type C MCBs. The board is supplied from a UPS output.

With the mains present and the system on bypass, the prospective fault current at the board might be high enough to operate the magnetic element of the MCB very quickly. The measured Zs may look acceptable.

Now consider the same fault during battery operation. The inverter can only deliver a limited output current. If that current does not reach the instantaneous trip threshold of the type C MCB, the breaker must rely on its thermal curve. The fault may last too long, or the UPS may shut down before the final-circuit breaker operates.

The result is not just a nuisance trip. It is a design conflict between the source capability and the protective device operating characteristic.

This is why the phrase “check the disconnection time on inverter” should appear in every technical review of a backed-up distribution system.

How to overcome low fault current problems

The solution is rarely one single change. A good design checks the source, the protective devices, the earthing arrangement, and the operating modes together. The aim is to make sure the right device trips quickly, even when the UPS is the only source.

Get the manufacturer’s fault current data early

Ask for the short-circuit performance of the UPS output before selecting downstream protection. The useful data is not just the rated kVA.

Request or confirm:

• maximum short-circuit current from the inverter

• duration of available fault current

• overload curve

• behaviour under line-to-earth faults

• bypass transfer behaviour during faults

• output earthing requirements

• recommended downstream protective devices

• selectivity guidance

  
  

For larger systems, ask for time-current curves or protection study data. If the manufacturer cannot provide enough information, treat the design risk as unresolved.

Design for the worst credible operating mode

Do not prove compliance only with the mains healthy. Check normal mode, battery mode, static bypass, and maintenance bypass where relevant.

The limiting case is often battery mode because available fault current is lowest. If the system complies in that condition, it is usually much easier to coordinate higher fault levels in bypass, although high fault current can introduce breaking capacity and selectivity issues of its own.

This is where BS 7671, disconnection times, backup, circuit, compliance, wiring regulations all meet in a practical design task. The installation must remain safe in the mode it is intended to operate in, not just during commissioning tests with the public supply present.

Use protective devices that operate at lower fault currents

A common fix is to review the downstream device type and rating.

For example:

Device choice	Why it may help	Design caution
Type B MCB instead of type C	Lower instantaneous operating current	Check inrush currents and nuisance tripping
Lower-rated final-circuit protection	Reduces required fault current for operation	Must still suit load current and cable rating
RCBO or RCD protection	Earth fault disconnection does not rely on high overcurrent	Check discrimination, leakage current, and UPS compatibility
Electronic trip MCCB	Adjustable settings can improve coordination	Requires correct setting, testing, and documentation
Semiconductor fuse or special fuse	May suit some power electronic systems	Must match manufacturer guidance

RCDs can be a strong answer for earth fault protection because their operation depends on residual current rather than high prospective fault current. That said, they are not a universal fix. UPS leakage current, filters, DC components, and earthing arrangements can affect RCD selection. Type AC devices are often unsuitable where electronic equipment can produce non-sinusoidal residual currents. The correct RCD type must be selected for the equipment and installation.

Testing should confirm the circuit conditions that calculations are based on.

Reduce circuit impedance where practical

Reducing impedance helps increase fault current, although it cannot overcome the inverter’s hard current limit. It can still make the difference on marginal circuits.

Practical measures include:

• keeping UPS-backed final circuits short

• using larger CPCs where needed

• avoiding unnecessary sub-distribution distance

• locating the UPS close to the backed-up board

• checking terminations and protective conductor continuity

• separating essential loads from general loads

  
  

Shorter, simpler circuits are easier to prove. Long final circuits supplied from a current-limited inverter are where disconnection time problems often appear.

Use distribution architecture that limits the risk

Avoid feeding a large general-purpose board from a small UPS and hoping protective devices will work as usual. It is usually better to create a dedicated backed-up board with carefully selected circuits.

Good practice includes:

• keep backed-up circuits limited and clearly labelled

• separate critical and non-critical loads

• use local final-circuit protection matched to inverter fault levels

• avoid spare ways being used later without design review

• document the UPS operating modes and protection assumptions

  
  

A future alteration can undo a safe design. Labelling and certification should make clear that the board is fed by a current-limited source.

Consider an output transformer or defined earthing arrangement

Some installations need a separately derived supply with a defined neutral-earth reference on the UPS output. An output transformer can help create a clear earthing arrangement and manage certain fault conditions, although it adds cost, losses, space, and another impedance into the system.

The decision should come from the protection strategy, not habit. A transformer can clarify the fault path, but it does not automatically solve low fault current. In some cases, the extra impedance reduces fault current further. The design still needs calculation.

Coordinate with the bypass supply

If the UPS can transfer to bypass during downstream faults, the bypass path may provide enough current to clear protective devices. That can be useful, but it must be verified.

Check:

• whether bypass is available during mains failure

• whether transfer is inhibited for some fault types

• whether the bypass protective device can clear the fault

• whether downstream devices remain selective on bypass

• whether breaking capacities are adequate for bypass fault level

  
  

A system may have low fault current on inverter and high fault current on bypass. Both conditions matter. The downstream device must survive and operate correctly in both.

Use protection studies for larger or critical installations

For larger systems, a formal protection coordination study is often the cleanest way to avoid guesswork. The study should include time-current curves for upstream devices, downstream devices, UPS inverter limits, bypass fault levels, and cable impedances.

The output should answer practical questions:

• Which device trips first for a final-circuit fault?

• Does it trip within the required time?

• What happens in battery mode?

• What happens on bypass?

• Is selectivity maintained for likely faults?

• Are device breaking capacities adequate?

  
  

This is especially important for healthcare, data, industrial controls, life safety interfaces, and sites where loss of supply creates significant operational risk.

A protection study should include inverter mode and bypass mode.

What to document at handover

The design is only useful if the next person can understand it. Handover information should include enough detail to prevent unsafe changes later.

Include:

• UPS make, model, rating, and operating modes

• inverter short-circuit current data used for design

• bypass fault level and upstream protection details

• earthing arrangement for each operating mode

• protective device types, ratings, and settings

• RCD types and ratings, where used

• calculated disconnection times or verification method

• limitations on adding circuits or changing breaker types

• test results taken with mains present, and any calculations that cover battery mode

  
  

Where a standard loop test cannot reproduce the inverter-limited fault condition, say so. A measured Zs value alone may not prove battery-mode disconnection. Calculations and manufacturer data may be needed to demonstrate compliance.

The main takeaway

Low fault current is an inherent feature of many inverter-supplied systems. It protects the power electronics, but it can stop conventional protective devices operating fast enough. The risk is highest when the installation is running on battery, because the normal mains fault level is no longer available.

The practical answer is to design the backed-up distribution around the real source characteristics. Get the manufacturer’s short-circuit data, check the worst operating mode, select protective devices that can operate at the available current, use RCDs where appropriate, keep circuits short, and document the assumptions clearly.

A backed-up supply is only safe if it disconnects under fault as well as it supports the load. Continuity and protection have to be designed together.

We use specialist design software to ensure full compliance for your UPS system. If in doubt, contact us and we'll be happy to help.