f97.beERIK SCHMIDT’S WORKSHOP
← All terms

Glossary · Energy industry & energy transition

Critical Load

Critical load critical load resilience emergency power security of supply Infrastructure

Critical Load refers to electrical loads that must continue to be supplied with priority during a fault, a power outage or planned island operation. In German, the usual terms are kritische Last, kritischer Verbraucher or, in planning practice, notstromversorgte Verbraucher (loads supplied by emergency power). What is meant is not a particularly high load, but a load with a particular function. Its failure would have immediate consequences for safety, health, controllability, data integrity, process stability or the restoration of supply.

The term must not be confused with critical loads in environmental science. There, the plural refers to limits of tolerable pollutant inputs into ecosystems, for example acid or nitrogen deposition. This entry deals exclusively with the electrical engineering meaning.

What counts as critical

Critical loads can be, for example, medical devices, safety lighting, fire alarm systems, access systems, control centres, radio and communication systems, servers, cooling, pumps, controls of industrial processes or parts of building services. The classification cannot be derived from the type of device alone. A pump in an office building may serve comfort only, whereas in a hospital or waterworks it may be essential for supply. A server that only holds archive data is dispensable in an emergency. But if dispatch planning, grid control or production safety run through it, it becomes a central load.

Criticality therefore arises from the context. Cooling becomes critical when it protects medicines, food, batteries or data centre equipment. A door control becomes critical when it affects escape routes, access to security areas or rescue operations. An inconspicuous network switch can determine the controllability of an entire installation. Conversely, large loads can be non-critical if their failure only affects comfort or production that can be postponed in the short term. A common mistake is to treat criticality as a fixed technical property of a device.

Power, energy and time

Technically, a critical load is described by several quantities. Power, usually stated in kilowatts, indicates how much electrical power the load requires at a given moment. The energy demand in kilowatt-hours results from power and supply duration. A simple list of devices is therefore not sufficient for sizing emergency generators, battery storage systems or uninterruptible power supplies (UPS). Information is needed on the load profile over time, on inrush and starting currents, on the permissible interruption time, on the minimum runtime and on priority. An X-ray device may briefly demand high power but runs only for a short time. A communication system may need little power but must run stably for many hours. For example, with a continuous load of 2 kW and a bridging period of 72 hours, the energy demand is 144 kWh, excluding efficiency losses and reserves.

The permissible interruption time is often the most stringent criterion. One example is medically used areas in Germany. The installation standard applied there, DIN VDE 0100-710, classifies loads by changeover time classes. Particularly important are the classes with a maximum of 0.5 seconds, for example for operating theatre lights and life-supporting devices, and with a maximum of 15 seconds for many further loads of the safety power supply. The time class describes a functional requirement but does not prescribe a specific technology. The entire chain must meet the deadline: detection of the voltage failure, start of the replacement source and load transfer. For 0.5 seconds, therefore, practically only battery-backed systems or a UPS come into question. For 15 seconds, a quickly starting generator set may suffice.

The supply duration must also be defined. A UPS typically bridges seconds to minutes until a generator has started up. A battery can cover several hours if the load is small enough. An island grid with its own generation can run longer if fuel, the storage state of charge, the generation profile and load management fit together. In its emergency power guideline, the Federal Office of Civil Protection and Disaster Assistance (BBK) recommends keeping fuel in stock for at least 72 hours of operation without resupply. Power quality also includes reactive power, starting currents and stable voltage, because some devices need not only energy but a stable electrical environment. A critical load is therefore always tied to a target: which function is to be maintained for how long under which fault conditions?

Distinction from related terms

  • Peak load: the highest power demand in a period under consideration. It can arise from comfort consumption, production peaks or the simultaneous use of many devices without being critical.

  • Base load: a permanently present power demand or, in the power plant context, a mode of operation. A critical load can occur permanently or only temporarily.

  • Critical infrastructure: facilities and services whose failure can have significant societal consequences. Critical Load, by contrast, refers within an installation or a section of a grid to the specific electrical loads that must be supplied for this function. Businesses outside critical infrastructure also have critical loads.

Planning: classify instead of backing up everything

The practical significance lies in planning the security of supply of individual buildings, sites, districts or installations. Hospitals, data centres, waterworks, traffic control systems or industrial plants need concepts that distinguish between dispensable, deferrable and absolutely essential loads. A classification into immediately indispensable, temporarily required, deferrable and disconnectable has proven effective. It determines which loads are connected to a UPS, which are backed up by diesel generators, batteries or fuel cells, and which may be automatically disconnected by load shedding in the event of shortages. Electrically, this is usually implemented via dedicated emergency-power circuits and sub-distribution boards.

This separation is particularly important for sizing batteries and for island operation. Anyone who bases the design on the entire connected load of a building quickly oversizes storage and emergency power systems. Anyone who defines the critical load too narrowly ends up with a system that works on paper but loses central functions in the event of a fault. Planning must clarify which loads have to run simultaneously, which can be staggered in time and which are critical only in certain operating states. Restart also belongs here: after longer interruptions, cooling, heating or pumps can draw considerably more power on reconnection than in normal operation. This effect is described by cold load pickup.

The BBK guideline on emergency power supply in companies and authorities was revised in October 2024 together with the standardisation organisation DKE. It describes a staged approach. First, it is determined how much power-dependent infrastructure exists and which areas are to be supplied by a generator set or battery storage. Then follow practical emergency operation, fuel storage and regular tests and exercises, since only these reveal weaknesses. Battery storage, fuel cells, renewable energies and local island grids are now taken into account.

Photovoltaics and storage are not emergency power without additional technology

With the energy transition, decentralised generation, storage and controllable loads are increasingly combined in local supply concepts. However, a rooftop photovoltaic system does not automatically supply critical loads during a grid outage. In Germany, a grid-parallel inverter must disconnect from the grid in the event of a grid outage. This is required by the grid and system protection (NA-Schutz) under the application rule VDE-AR-N 4105 for generating plants on the low-voltage grid. The reason is not only that such an inverter needs the grid's voltage and frequency as a reference. Above all, unintended feed-back into a de-energised grid is to be prevented, because it would endanger fitters working there.

If a PV system and battery are to supply critical loads, a backup-power-capable or island-capable inverter and a standard-compliant changeover and disconnection device are required. Under VDE-AR-N 4105, during backup power operation the connection to the public grid must be disconnected on all poles, that is, including the neutral conductor. In addition, a source must set frequency and voltage, and the protection technology must be designed for island operation. In island operation, inverters often supply considerably lower short-circuit currents than the grid. The protection concept must take this into account. Local generation alone does not replace an emergency power concept.

Responsibilities and legal framework

Grid operators are responsible for the operation of the public grid, not for the internal prioritisation of loads within a building. Operators of facilities must themselves determine which loads are critical in their case and how these are electrically separated, protected and switched. Planners and installers translate these requirements into switchgear, protection concepts, controls and testing processes. For certain classes of installations, standards set binding minimum requirements, for example DIN VDE 0100-710 for medically used areas. Sector-specific rules take precedence over general guidelines. They do not replace a site-specific load analysis.

In Germany, the KRITIS Umbrella Act (KRITIS-Dachgesetz) entered into force on 17 March 2026. It implements EU Directive 2022/2557 on the resilience of critical entities (CER Directive) and requires operators of critical installations, among other things, to carry out risk analyses and prepare resilience plans. A power outage is among the hazards to be considered. Which installations are specifically covered is to be determined by a statutory ordinance. According to information from professional practice, this ordinance was most recently still not available, so the operator obligations have so far applied only to a limited extent. For those affected, it follows that critical loads must not be identified only in a crisis. They must be documented in advance, tested and organisationally owned.

Economic trade-offs and resilience

Every load classified as critical causes costs: larger batteries, more powerful generators, redundant cabling, maintenance, fuel reserves, testing and higher demands on building automation. A definition that is too broad pushes normal comfort and operating functions into the emergency power area and makes the installation more expensive. A definition that is too narrow saves investment but increases the risk of failure. The classification is therefore not a purely technical question. It combines risk analysis, operating costs, liability and regulatory requirements.

The term also shows that resilience is not the same as maximum self-sufficiency. A site often becomes more resilient simply by reliably backing up a few central functions, rather than replicating the entire normal operation for rare fault events. Critical Load thus describes not a quantity of electricity but an operational priority: a load whose supply is to be maintained for a specific function under defined fault conditions. Those who use the term properly separate power from energy, criticality from size and security of supply from mere connected load.

Sources

Related glossary entries

Edited by eCMS on 11.11.2026 18:17

© 2026 Erik Schmidt. All rights reserved. Quotations with a link to the source are welcome.