Connection capacity

In an electrical context, connection capacity refers to the power demand of connected devices, the technical design of a connection, or the power limit agreed for a specific grid connection. These meanings must be kept apart. A cable of sufficient size does not in itself give rise to a right to draw any technically transmissible power from the public grid. Conversely, the agreed capacity says nothing about how heavily the connection is actually used. For installations with both power consumption and feed-in, the two directions must be considered separately.

Connection capacity is a power quantity, not an annual amount of energy. Active power is stated in kilowatts (kW) or megawatts (MW), apparent power in kilovolt-amperes (kVA) or megavolt-amperes (MVA). Which quantity is meant must be clear from the documents and the context. For a device connection, the rated power of the device may also be meant. This entry mainly deals with electrical grid connections; the term is also used for gas or heat connections, but with different technical constraints.

Distinguishing power, energy and connected load

Electricity consumption describes an amount of energy over a period of time and is stated, for example, in kilowatt-hours (kWh). A consumer drawing a constant 2 kW needs 6 kWh in three hours. A consumer with 6 kW can convert the same amount of energy in one hour. The second case places a greater load on the connection, even though both need the same amount of energy. For cables and other equipment, current, load duration and thermal conditions are decisive in particular, not annual energy alone.

A deliberately simplified example illustrates the orders of magnitude: A 100 kW installation that runs only ten hours a year needs 1,000 kWh. By contrast, a 1 kW installation in continuous operation all year needs 8,760 kWh in a year of 8,760 hours. The first thus has the lower energy demand but a hundred times the operating power. A low annual consumption therefore does not allow the conclusion that the required connection capacity is low.

Installed capacity is also not automatically the grid connection capacity. Suppose a photovoltaic system has 15 kWp of module capacity and an inverter with an active power output of at most 10 kW: The module rating describes the direct-current power under defined test conditions, not the feed-in to the grid at any given time. If the building simultaneously consumes 3 kW, an inverter output of 10 kW leaves a calculated 7 kW of feed-in at the connection. An agreed feed-in limit must nevertheless be observed regardless of the self-consumption that happens to be present.

The example also shows why a single figure without a reference point says little. For a heat pump, electrical power input and heat output delivered must be distinguished. For a battery storage system, charging and discharging power in kW must be separated from the storable amount of energy in kWh. For the grid connection, what counts in each case is which electrical power can actually flow across the connection point under the permissible operating conditions.

What fuses and power ratings tell you

The fuse protection of a connection is stated in amperes. For a symmetrically loaded three-phase system: S = √3 × U × I. Here U is the voltage between two phase conductors and I is the current per phase conductor. At 400 V and 63 A, this gives a calculated apparent power of about 43.6 kVA. Active power follows from P = S × λ, where λ is the power factor. At λ = 1, 43.6 kVA thus corresponds to 43.6 kW; at λ = 0.9, to approximately 39.3 kW.

This calculation is a physical example, not a blanket release for a service connection fused accordingly. With uneven distribution across the phase conductors, a single phase may already be fully loaded even though the total power is, on paper, still below the limit. A fuse is also not a power limiter set to an exact kW value. For the permissible use in practice, the connection agreement, cables, protective devices and meter position must fit together.

Long-lasting loads in particular, such as from charging or electric heating, require a corresponding design. VDE FNN points out that the load capacity of a meter position depends on the operating mode and wiring. For Germany, the TAR low voltage forms the basis of the grid operator-specific technical connection conditions (TAB). The publicly accessible VDE page names VDE-AR-N 4100:2026-04 as the valid version. However, no blanket obligation to retrofit every existing installation can be derived from this.

Simultaneity and load management

For planning, simultaneity is decisive: Which consumers or generators reach which power at the same time? Adding up all nameplate values can overestimate the demand. An unjustifiably low simultaneity assumption, on the other hand, can underestimate it. The appropriate assessment depends on usage, the number of installations, operating procedures and, where applicable, a reliably enforced power limitation. Power management can reduce the peak occurring simultaneously.

Four charging points of 22 kW each have a combined installed charging capacity of 88 kW. If a load management system reliably limits their joint power consumption to 44 kW, at most 44 kW is available to them together. This does not mean, however, that a 44 kW connection is sufficient for the entire property: Further consumers and technical constraints must also be taken into account. A statistical expectation that all vehicles rarely charge at the same time is not the same as a technically safeguarded limitation.

Another calculation example: If 20 kW is available for charging uninterrupted over eight hours, at most 160 kWh can be drawn from the grid through it. Charging losses reduce the energy that arrives in the vehicle batteries. Whether this is sufficient depends on the actual recharging need and the departure times. Charging spread over time therefore does not automatically reduce the energy demand, but can meet the same demand with a smaller power peak.

Dynamic load management takes into account the other consumption at a site and distributes the remaining power in each case among controllable applications. Static load management, by contrast, works with a fixed upper limit for the group of installations concerned. Planning and operation must also take fault cases into account: A communication failure must not cause previously limited devices to draw their full power in an uncoordinated manner.

A storage system shifts the limits but does not eliminate them. If it is to fully cover an additional load of 10 kW for two hours, it needs, in idealised terms, 20 kWh of usable energy and at least 10 kW of discharge power; losses come on top. An empty storage system cannot take on this task. Likewise, solar power is not a reserve available at any time for a winter consumption peak. Such contributions may only be assumed in line with their assured availability.

Significance for grid planning and connection capacity

The requested capacity is an important starting variable of the grid compatibility assessment, but does not replace the technical investigation. Among other things, permissible currents and voltages, grid disturbances and the safe functioning of the protection technology must be considered. Connection capacity must not be confused with short-circuit power: The latter characterises other electrical properties of the grid and is not a power available for drawing in normal operation.

In the distribution grid, individual connections add up to a joint load on cables and transformers. Two sites with a consumption peak of 50 kW each load a jointly upstream grid section differently: If the peaks occur simultaneously, up to 100 kW results there from both together; if they occur one after the other, the contribution can remain significantly smaller. Annual energy amounts alone do not show this difference.

Similar usage patterns can increase simultaneity. This applies, for example, to electric heating in cold weather, identically oriented solar installations in sunshine, or many charging processes during a shared favourable price window. A mode of operation that is economical for the individual customer is therefore not automatically favourable for every local grid section. Connection capacity describes the framework; the load and feed-in profile over time determines how much that framework is actually stressed.

Connection capacity or grid capacity, by contrast, often refers to the possibility of providing or absorbing additional power at a particular location. The use of the term is not uniform everywhere. In any case, a rarely used existing connection is not a freely available reserve for any new connections. For this, the existing usage rights, the joint grid load and the technical limits must be taken into account. Conversely, the sum of all agreed connection capacities is not to be equated with a grid load that actually occurs permanently.

Contract, costs and German connection rules

Even for small low-voltage connections, the capacity to be made available is not merely a fuse rating: § 4 NAV counts it among the contract particulars. Under § 19 paragraph 2 NAV, capacity-relevant extensions must be notified under the conditions stated there. Charging equipment must be registered before commissioning; if its total rated power per electrical installation exceeds 12 kVA, the prior consent of the grid operator is additionally required. This threshold must be distinguished from the rules for controllable consumption devices.

Within the scope of the NAV, a construction cost contribution under § 11 may only relate to the part of the capacity requirement above 30 kW. It must be separated from the costs of the actual grid connection. This implies neither a free connection up to 30 kW nor a general 30 kW exemption threshold for medium-voltage, storage or generation connections. The legal basis, type of connection and specific agreement must each be examined separately.

For a business's site decision, more therefore counts than the electricity price. If, for example, a production process needs its power at fixed times, power that is available only intermittently may be insufficient. A charging operation with long standing times may be able to cope well with the same restriction. What should be compared economically are connection expansion, operational adaptation, control and, where applicable, storage – in each case under the same requirements for availability and throughput.

Controllable and flexible use

In Germany, special rules apply under § 14a EnWG to defined controllable consumption devices in the low-voltage grid. In principle, installations put into operation since 1 January 2024 with more than 4.2 kW of grid connection capacity are covered: non-public charging equipment, heat pumps including supplementary heaters, space cooling, and electricity storage systems with regard to their grid consumption. Aggregation rules apply to certain groups of installations; transitional and exemption provisions apply to existing installations.

With a permissible grid-oriented control, grid consumption is temporarily limited; in return, operators receive reduced grid charges. The minimum power in the case of direct control is in principle 4.2 kW per device, with special rules for large heat pumps, for example, and for control via an energy management system. Normal household consumption remains unaffected. This limitation is neither a blanket 4.2 kW limit for the entire service connection nor a rule limiting its feed-in.

In addition, § 17 paragraph 2b EnWG enables flexible grid connection agreements with static or dynamic limits for withdrawal or feed-in. In particular, the power level, limitation periods, contract term, technical implementation and liability for exceedances must be specified. Special provisions under § 8a EEG and § 14a EnWG remain unaffected. Where a capacity is promised, it must therefore be evident whether and under what conditions the stated capacity can actually be used.

Limits of the indicator

A high connection capacity is no promise of uninterrupted supply. It describes neither the frequency of grid disturbances nor the availability of a backup supply. Likewise, a large storage system does not automatically mean that it can supply the site during a grid outage. Additional technical properties would be required for that. Connection capacity, security of supply and operational resilience to outages answer different questions.

For robust planning, at least five details should therefore be established: reference point, direction of power flow, active or apparent power, permissible use over time, and the method for complying with the limit. Realistic operating profiles and expansion needs come on top. Only then does an isolated kW or kVA figure become a usable planning variable. Connection capacity thus links devices and buildings with the technical, contractual and economic conditions of their grid use.

Sources

Niederspannungsanschlussverordnung – NAV — Bundesministerium der Justiz und Bundesamt für Justiz; unbekannt. § 4 Absatz 1 Nummer 4; § 11 Absätze 1–5; § 19 Absatz 2
Technischer Leitfaden Ladeinfrastruktur Elektromobilität, Version 4.1 — DKE, VDE FNN, BDEW, ZVEH, ZVEI und VDA; 2023-08. Abschnitt 3.1 „Bedarf an Anschlussleistung“, insbesondere Abschnitte 3.1.1 und 3.1.2
Technische Anschlussregeln Niederspannung – VDE-AR-N 4100 — VDE FNN; 2026-05-13. Abschnitte „Inhalte“, „Neuerungen der gültigen Fassung“, Erläuterung der TAB sowie FAQ „Strombelastbarkeit Unterverteiler“
Integration von steuerbaren Verbrauchseinrichtungen — Bundesnetzagentur; unbekannt. Abschnitte zu betroffenen Anlagen, Gültigkeit, Mindestleistung, Haushaltsverbrauch und Netzentgeltreduzierung
Energiewirtschaftsgesetz – § 17 Netzanschluss — Bundesministerium der Justiz und Bundesamt für Justiz; unbekannt. Absatz 2b

Related glossary entries

Edited by eCMS on 11.11.2026 13:50