Can a BESS keep an entire factory running during a power outage?

The question comes up in almost every conversation about industrial battery storage and production security: can the battery keep the whole factory running if the power disappears? It is understandable, but also the wrong question. The answer does not depend primarily on whether a BESS (battery energy storage system) is large enough, but on which parts of production actually need to stay online, for how long, and how quickly any backup must be able to take over.

This article builds on the technology differences between BESS, UPS, generator, and STS covered elsewhere, and takes the next step: how to translate a battery's capacity into a realistic runtime for a given load, and why "the whole factory" is rarely the right design goal.

Why the question is framed incorrectly

"Can the battery keep the whole factory running?" sounds like a yes/no question about technical capacity. In practice it mixes several questions together, and without answers to them, no system, regardless of size, can be said to be "large enough." Guidance from Energistyrelsen (the Danish Energy Agency) on emergency power for businesses makes the same point from the other direction: "It is not a given that it is necessary to be able to maintain all of an organisation's functions in the event of a power outage." Each organisation must weigh which functions actually need to continue, and for how long, instead of automatically assuming that everything must be covered. The same applies to a manufacturing business: the question is not whether a system can cover everything, but what actually needs to be covered.

The better questions

Instead, there are four questions that lead to a usable answer:

1. Which parts of production must never go down, which is rarely the entire factory, but typically a bounded group of equipment and processes?
2. How long must they stay online, where a one-minute interruption is a completely different task from an hour or a day?
3. How quickly must backup take over, because some loads such as PLC controls can fail at even a short voltage deviation, while others can tolerate a longer interruption without consequence?
4. And what does each minute of stoppage cost, which ultimately determines whether the investment in backup makes sense, a calculation covered in detail elsewhere in this article series?

Energistyrelsen's guidance frames the second and third questions as a scale: 1 second, 1 minute, 10 minutes, 1 hour, 2 hours, 24 hours, or 1 week, and notes that ventilation, cooling, elevators, and pumps can often tolerate an interruption of around one minute with limited consequence, while other functions can be highly critical even at shorter interruptions. The requirement for response speed is therefore not the same across the whole factory; it varies load by load.

The whole factory versus critical loads

Once the four questions are answered, the difference between "the whole factory" and "critical loads" becomes a pure sizing question. Keeping the factory's full power consumption running requires a system that matches the full power demand for the full runtime, while selected critical loads only require a system sized to that much smaller subset. NREL's fact sheet on battery containers and resilience describes exactly that approach as standard practice: a critical load is typically estimated as a smaller subset of total consumption, either as a percentage of normal consumption, via a dedicated critical-load panel with its own metering, or by summing the power demand of the individual components, and not all critical loads are equally critical: some need power the moment the grid drops out, others can tolerate a longer interruption.

A system sized for the whole factory is markedly more expensive than a system sized for the loads that are actually critical, without necessarily delivering more real security. If the consequence of briefly losing non-critical production is manageable, while the consequence of losing cooling, controls, or safety systems is serious, it is rarely a sensible priority to spend capacity on the former rather than securing the latter.

Which loads are actually critical varies from business to business and always requires a concrete review of the specific production, but across industrial environments some categories often recur:

• Cooling and freezing: a temperature swing over minutes to hours can ruin an entire batch of raw materials or finished goods.
• Pumps: for process cooling, water supply, or wastewater, where a stop can create knock-on damage further down the process.
• PLC and SCADA: control and monitoring of production equipment, which can enter a fault state from a short voltage swing even if the machine physically could have managed without power.
• Servers and IT: systems that support production control, quality data, or traceability.
• Process control: ongoing regulation of temperature, pressure, or flow, where an interrupted sequence can cost the entire batch.
• Safety systems: emergency lighting, alarms, and access control, often critical for reasons other than economics.
• Selected production lines: rather than the entire plant, if those lines carry the most valuable or vulnerable part of production.

The list is not exhaustive. It is a starting point for the conversation every business needs to have internally before it is possible to talk meaningfully about what a backup system should actually be designed to cover.

Power versus capacity: the decisive technical distinction

Once it is clear which loads are critical, the next question arises: can a given BESS actually supply them? The answer requires keeping two independent properties of the system apart. Power, measured in kW (kilowatts), determines how much can be supplied at once, the maximum load the system can deliver at a given moment. Capacity, measured in kWh (kilowatt-hours), determines how long that power can be sustained, the total energy available before the battery is empty. According to the IEEE Smart Grid Initiative's review of battery container properties, the suitability of a given storage technology for a given task is primarily determined by exactly these two independent quantities: its power and energy capacity, and the rate at which they can be charged and discharged.

A system with high power but low capacity can deliver a lot of power at once, but only for a short time. A system with lower power but high capacity can supply a smaller load for markedly longer. That is precisely why "can the battery keep the whole factory running" cannot be answered from a single number: two numbers are required, the load size in kW and the desired duration, which together determine the required kWh need.

From module figures to concrete runtime

A standard module in an industrial BESS typically has a capacity of about 261 kWh and a power rating of about 125 kW. The figures do not automatically mean anything concrete until they are translated into runtime for a specific load, because 261 kWh is only an abstract energy number until it is held against how much power the load it is meant to supply actually draws.

The basic relationship is simple: runtime equals available capacity in kWh divided by the critical load's power in kW. But the raw calculation overestimates real runtime, because it does not account for three factors that pull toward less available time than the nominal figure suggests. Efficiency means there is always an energy loss when energy is converted and transferred through the system's electronics and cables, so the energy that reaches the load is lower than what is stored in the battery cells. Safety margin means a sensible design does not plan to use capacity all the way to zero, partly as a buffer against unforeseen conditions, and partly because very deep discharge is not good practice for battery life. Reserved SoC (state of charge) means that if part of the capacity is earmarked for other purposes, such as commercial optimisation, or reserved as a floor the system does not go below, it is not available for the backup event.

An illustrative, simplified calculation

To make the relationship concrete, not as an operational guarantee, but as a pure calculation exercise to illustrate the method: assume a module of 261 kWh nominal capacity that is to supply a hypothetical critical load of 40 kW, consisting for example of cooling, pumps, and PLC/SCADA control.

Simple division of 261 kWh by 40 kW gives a little over 6.5 hours of runtime, but that figure is the theoretical ceiling, not a realistic estimate. Assuming, solely as illustrative, round example figures and not as documented ZynexGroup values, that about 10 percent of the energy is lost in conversion and transfer, that a safety margin of 15 percent of capacity is reserved, and that a further 10 percentage points are reserved for other purposes, the actually available capacity becomes around 65 percent of the 261 kWh, about 170 kWh. Divided by the same 40 kW load, that gives an illustrative runtime of a little over four hours instead of the theoretical 6.5. The difference between the two figures is the point itself: a nominal kWh figure is not the same as the actually available runtime for a given load.

The above is a simplified, illustrative calculation with round assumptions, not a guarantee of runtime on any concrete installation. Actual runtime at a given location depends on the system's documented SoH (state of health), the concrete configuration, and conditions at the individual site. Battery cells lose capacity over time and with use, which is documented broadly in the peer-reviewed literature on lithium-ion battery degradation, and a system's real usable capacity must always be assessed from documented SoH, not from the original nameplate figure.

Why backup requires separate technical design

Even when runtime and critical load are calculated correctly, the calculation is only half the task. That a system mathematically has enough capacity to supply a critical load does not mean it can automatically do so in practice when the grid drops out. The system must detect the outage, disconnect safely from the failed grid, and supply the prioritised load isolated from the rest of the installation, typically via a dedicated transfer switch such as an STS or equivalent. NREL's material on battery containers for resilience emphasises that a grid-connected battery system can only deliver power during an outage if it is installed with the right islanding controls in place, not as an automatic property of having sufficient capacity; the same design logic is covered in detail on STS technology elsewhere in this article series.

A standard BESS therefore does not automatically function as emergency power or UPS for any load, neither the whole factory nor a single critical machine, unless it is deliberately designed and configured for that. Backup is always an optional process that requires a concrete analysis of the business's electrical installation, critical loads, and risk level, plus a separate technical design.

ZynexGroup's approach to system assessment

In our standard model, we optimise an industrial BESS's flexibility commercially in the market, against a fixed contractual availability payment to the customer. The standard model does not automatically include emergency power or backup, neither for the whole factory nor for selected critical loads. If a customer wants to explore whether reserved capacity or a backup solution makes sense, we start with the same exercise as this article: identify the loads that are actually critical and calculate their realistic runtime from the actual, adjusted kWh figure, not the system's nominal capacity. If there is a concrete reason to go further, we coordinate the process and work with partners on the technical design and the installation itself on site.

Overall perspective

"Can a BESS keep the whole factory running?" is not the question that leads to a usable answer. The questions that do are: which parts of production must never go down, how long must they be able to stay online, how quickly must backup take over, and what does each minute of stoppage cost? Once those are answered, it becomes clear that critical loads can often be secured far more economically than the whole factory, because the required capacity in kWh is a function of the actual load in kW, not of the factory's size. A module with about 261 kWh capacity and 125 kW power is only a meaningful figure when it is translated into runtime for a concrete load, adjusted for efficiency, safety margin, and reserved capacity, and it always requires a separate technical design to make that capacity actually available as backup.

Assessing the potential for your facility

If you are considering whether reserved capacity or a backup solution makes sense for your production, the natural starting point is a mapping of your critical loads and their actual power demand, together with data on your grid connection and consumption profile. We use that kind of data for an initial assessment of whether there is a match between your situation and our model, and whether there is a concrete reason to proceed with a technical analysis. You are welcome to contact us here.