Hosting Capacity · Glossary

Hosting capacity refers to the ability of an electricity grid to accommodate additional generation, storage or loads without violating technical limits and without requiring grid reinforcements or changes to control. It usually means the additional connectable capacity for a grid section, a grid connection point or a voltage level. This may be, for example, output from photovoltaic systems or capacity drawn by heat pumps, wallboxes, battery storage systems or new industrial facilities. The US research laboratory NREL defines the term similarly for distributed photovoltaics. However, it stresses that hosting capacity is not a hard upper limit: it increases with every grid reinforcement.

The term is used mainly in the distribution grid. Many new connection requests arrive there at the same time. However, these grids were not originally built for solar power to be fed back everywhere, for many cars to charge at the same time or for buildings to be heated electrically. Hosting capacity is therefore not an abstract expansion reserve. It describes a concrete technical capability to accommodate additional capacity, which depends on assumptions about feed-in, consumption, simultaneity and grid operation.

Technical significance

Hosting capacity is usually stated in kilowatts or megawatts, and in studies sometimes also as a share of the load or of the transformer rating. Such a figure is only meaningful if it is clear what it refers to. It can apply to a single house connection, a low-voltage feeder, a local distribution transformer, a medium-voltage line or a substation. It can concern additional feed-in, additional load or both. For photovoltaics, the additional feed-in capacity usually matters; for electromobility and heat pumps, the additional demand capacity. The capacity to accommodate feed-in and the capacity to accommodate load are usually of different size.

Which limit is reached first depends on the case. For photovoltaics in the low-voltage grid, it is often voltage regulation: if many systems feed in at the same time while little is consumed locally, the voltage rises at the end of a feeder. The benchmark for this is the standard DIN EN 50160. In normal operation, it permits slow voltage variations in the low-voltage grid of about 230 volts plus or minus 10 percent. The VDE application rules for grid connection additionally limit how much generation plants may change the voltage. This leaves room for load fluctuations and for the upstream medium-voltage grid. For new loads, the thermal loading capacity of cables, overhead lines or transformers is often the limit. Other possible limits are protection concepts, short-circuit power, power quality, reverse power flow into higher voltage levels and operational reserves.

A hosting capacity figure therefore always depends on the model. The calculation takes into account grid data, load and feed-in profiles, assumed simultaneities, permissible voltage bands, equipment data and safety criteria. Location also plays a role: the same system is often uncritical close to the transformer but possibly not at the end of a long feeder. NREL stresses that the hosting capacity of a line cannot be described by a single number. Rather, it marks the point from which reinforcement or changed control may become necessary. If the assumptions change, the result changes too.

Distinction from grid connection capacity and grid expansion

Hosting capacity is not the same as contractually agreed grid connection capacity. The latter determines what capacity a particular connection user may draw or feed in. Hosting capacity, by contrast, describes what an entire grid area can technically accommodate before additional measures become necessary. A grid section may still have headroom in calculations while an individual connection is limited for other reasons, for example because of a house connection cable that is too weak. Conversely, a connection may be possible even though the grid first has to be modified for it.

The term should also not be equated with grid expansion. A low hosting capacity does not mean that a grid is permanently unsuitable. It initially only shows that, under the assumptions chosen, technical limits are reached. One can respond with conventional expansion, for example stronger cables, a larger transformer or additional switching options. Operating the grid differently is also possible: controllable local distribution transformers, reactive power control of inverters, controlled curtailment of generation, load control, grid-serving operation of storage or flexible grid connection agreements. All of this can increase the capacity that can be accommodated without every line having to be reinforced immediately.

Likewise, hosting capacity must be distinguished from flexibility. Flexibility is the ability of installations to shift feed-in or consumption in time. Hosting capacity describes how much connection and operation is possible under certain grid conditions. Flexibility can increase hosting capacity, but only if it is reliably available, technically controllable and legally usable. Without clear rules on control, remuneration and responsibility, the relief remains theoretical.

Legal instruments in Germany

Germany now has several rules that allow existing capacity to be used better. They do not change physics, but they change the assumptions under which a grid operator may calculate. The most important examples (as of November 2026):

Controllable consumption devices under § 14a EnWG: Since 1 January 2024, the determinations of the Bundesnetzagentur have applied to new wallboxes, heat pumps, air conditioners and electricity storage systems that are connected at low voltage and have a grid connection capacity of more than 4.2 kW. If overload threatens, the grid operator may temporarily curtail their power consumption, but in principle not below 4.2 kW. The rule concerns only consumption from the grid, not feed-in.
Flexible grid connection agreements under § 8a EEG: Since 25 February 2025, grid operators and operators of renewable energy installations can agree that the maximum feed-in capacity at the connection is limited, also only temporarily or at different levels depending on the time window. A grid connection point can thus be assigned more installed capacity than it could continuously accommodate. The wording does not impose any obligation to conclude such an agreement. In certain cases, however, the grid operator must examine this option and communicate the result. A parallel provision for withdrawal and storage exists in § 17 paragraph 2b EnWG.
In Germany, curtailment in the event of grid congestion has no longer been called feed-in management (Einspeisemanagement) since October 2021. It has been incorporated into Redispatch 2.0. Anyone who uses the term feed-in management therefore usually means either the old legal situation or, in general, the controlled limiting of feed-in.

These instruments show that hosting capacity is not a fixed value. If a grid operator calculates with the possibility of controlling installations, it may be able to permit more connections than if it has to assume that all installations run at full power at all times. Which instruments apply in an individual case depends on the type of installation, the voltage level, the commissioning date and the requirements of the grid operator.

Why hosting capacity matters in the distribution grid

With the energy transition, many investment decisions are made in the distribution grid. Photovoltaics are built on rooftops and open land, heat pumps replace boilers, electric cars charge in residential areas, and battery storage systems are located at the customer behind the meter or directly at the grid. Each individual installation is small compared with a power plant. In total, however, they considerably change load flows, voltage regulation and simultaneities in local grids.

Hosting capacity shows where these changes can be absorbed easily and where bottlenecks arise. For grid operators, it is a planning tool. For municipalities, project developers and connection users, it is an indication of which sites can probably be connected quickly and where reviews, waiting times or reinforcements are to be expected. In the USA, some utilities publish hosting capacity maps for this purpose in order to improve transparency for project developers. For regulation and policy, the term makes clear that the connection of new installations depends not only on installed capacity but on conditions in the local grid.

Spatial resolution is particularly important. A district may have enough transformer capacity overall while individual local grids can hardly accommodate any more photovoltaics. A city may have a robust medium-voltage grid while, in certain streets, the simultaneous charging of many vehicles brings the local distribution transformer to its limit. Averages conceal such bottlenecks. Hosting capacity is therefore practically useful only if it is considered at a sufficiently small spatial scale.

Typical misunderstandings

A widespread misunderstanding is to read hosting capacity as a fixed amount of free grid capacity, as if the grid were a storage room into which a certain amount of power still fits. Electricity grids work differently. How heavily they are loaded depends on the time, on the direction of power flow, on the spatial distribution of installations and on the behaviour of other grid users. A photovoltaic system can be critical for the grid at midday on a sunny spring day with little consumption; in the evening it does not feed in at all. A wallbox is unproblematic if it charges in a controlled manner. It becomes relevant when many vehicles charge simultaneously at high power.

A second misunderstanding is the statement that a grid is full. Technically, a grid section can be at full capacity in certain operating situations. It does not follow, however, that no further connection is possible. Conceivable options include a connection with a power limit, temporary curtailment, adapted protection technology, grid reinforcement or a different grid connection point. Talk of a full grid obscures which limit has specifically been reached and with which measure it could be remedied.

A third misunderstanding is to equate high hosting capacity with good grid planning. A grid that maintains very high reserves everywhere would be expensive, and the capital costs would be spread across all grid users. Grid planning has to weigh expansion, operational management, connection entitlements, security of supply and costs. A low hosting capacity can indicate a need to catch up. But it can also belong to a grid that was previously dimensioned economically and now has to fulfil new tasks.

Dynamic hosting capacity

Traditionally, hosting capacity is calculated statically: the grid operator checks whether an additional connection is permissible even in the worst case, for example with full feed-in and minimum load. More measurement data, digital grid management and controllable installations enable a more dynamic view. The question is then not only what capacity can always be connected without restriction, but what capacity can be used in which operating situations. NREL, for example, investigates how much additional distributed photovoltaics can be integrated with different control concepts.

This is particularly valuable with fluctuating feed-in and controllable loads. If grid operators know more precisely the voltages and the utilisation of their equipment, they can permit connection capacity that has to be limited only in a few hours of the year. Photovoltaic systems would then be curtailed during rare bottlenecks instead of being connected at a smaller size from the outset. Charging points could reduce their power when a local distribution transformer is heavily loaded. Storage could operate in a grid-serving manner if the rules for this are clear. Grid operators can thus resolve a grid congestion in a targeted way instead of dimensioning the entire infrastructure for rare peaks.

Technical possibility alone is not enough, however. Dynamic use requires metering infrastructure, communication technology, defined control rights, transparent connection conditions and a rule on how costs and benefits are treated. Conflicts arise where technical possibility, market rules and responsibilities diverge. If a battery storage system charges because electricity is currently cheap, for example, it can aggravate a local bottleneck. If it is to operate in a grid-serving manner, incentives and responsibility must be distributed differently. A contract model that lays down such conditions is the Flexible Connection Agreement.

Relationship to the market, connection law and system costs

Hosting capacity links technical grid questions with economic decisions. Project developers also assess sites by how quickly and at what cost a connection is possible. Grid operators have to check whether a connection can be established without violating the technical rules and what measures are needed for it. Regulation decides which investments are recognised, how grid charges work and whether flexible alternatives to expansion can be used economically.

For the system costs of electricity supply, this distinction is important. If every new installation is connected as if its full capacity had to be transported at all times, grid costs rise sharply. If one relies too heavily on curtailment and control, generation potential remains unused, or the rights of grid users become unclear. NREL therefore investigates not only when hosting capacity is reached but also what different ways of increasing it cost. Hosting capacity helps to conduct this trade-off concretely: Which limit is technically relevant? How often is it reached? Which measure lowers costs at the same security of supply? And who may decide on the operation of controllable installations?

Hosting capacity is thus more than a technical term of grid calculation. The term marks the interface between the physics in the local grid, connection processes, investment planning and the rules by which costs are distributed and flexibility is used. Used correctly, it does not say whether the energy transition fails because of the grid. It shows where additional capacity can be accommodated without adaptation, where limits arise and which decisions are needed to extract more usable hosting capacity from the existing infrastructure.

Sources

Advanced Hosting Capacity Analysis — National Renewable Energy Laboratory (NREL); unbekannt. Abschnitt What Is Hosting Capacity?
NREL's dynamic hosting capacity analysis (Fact Sheet) — National Renewable Energy Laboratory (NREL); unbekannt (Dokumentnummer FY19). Einleitungstext des Fact Sheets
Erneuerbare-Energien-Gesetz (EEG 2023) § 8a Flexible Netzanschlussvereinbarungen — Bundesministerium der Justiz / Bundesamt für Justiz (gesetze-im-internet.de); unbekannt (Vorschrift eingefügt mit Wirkung vom 25.02.2025). § 8a Absatz 1 bis 3 EEG 2023
§ 14a EnWG Steuerbare Verbrauchseinrichtungen – Integration von steuerbaren Verbrauchseinrichtungen — Bundesnetzagentur; unbekannt. FAQ-Seite für Verbraucher zu § 14a EnWG
Leitfaden für eine Optimierung der Versorgungsqualität, Version 1.0 — VDE FNN; unbekannt. Abschnitt zu den Merkmalen der Versorgungsspannung nach DIN EN 50160 (Tabelle Niederspannung)

Related glossary entries

Edited by eCMS on 12.11.2026 05:06