Proposal:Power circuits routing

Power circuits routing
Proposal status:	Draft (under way)
Proposed by:	fanfouer
Tagging:	power=circuit Data item
Applies to:
Definition:	Consistent tagging model for power circuits going along physical power infrastructure
Statistics:
Draft started:	2025-03-21

Power routing aims to document how power flows over a physical network, mainly between actual substations and along power lines. This proposal has been written in 2025 and largely based upon Bahnpirat and Surly work in the original power routing proposal. Several plea for finalizing what could be immediately implemented remain stuck. This is an attempt to push forward what could be achieved for now.

This presentation made on SOTM-EU 2023 intends to bring a summary of what is actually proposed there and what actually matters.

Proposal

It is proposed to approve relations with type=power + power=circuit, involving power lines (in a global sense, including lines cables and minor lines) and substations. They represent the continuous actual paths along which power flows through physical grids.

Another power relation would be useful to describe power line sections, with type=power + power=line_section. It is a subpart of an actual circuit, composed of several line segments (otherwise, the only segment is the section itself without requiring to involve it in a singleton relation).

It is also proposed to review two new physical properties for both lines and circuits as to allow proper power modelisation of power flow.

• resistance=* a numerical value quantifying the ability for a conductor to oppose to a continuous electrical current
• reactance=* a numerical value quantifying the ability for a conductor to oppose to a alternative electrical current

Rationale

As OpenStreetMap represent power lines in a synthetic way, often aggregating different lines sharing the same towers on the same way, the power lines won't reflect the actual flow of power. It is nevertheless possible to use the relational model to assemble different line sections to compose actual power paths just like it's done for public transportation over highways ways.

Such mapping is an established practice since 2017.

This proposal isn't intended to overload mapping practice with relations. Its authors and many people involved in the seek of describing power circuits had been concerned by finding the most neat solution as possible. However, some situations require to use relation as to provide a robust description of logical concepts over physical assets of power grids.

Power circuits

A power circuit is defined by 601-02-28 IEC60050 definition. It consist in the most extended part of a power grid between circuit breakers, mostly located in substations. It is composed of as many power line sections as necessary.

In practice, we should find continuous conductors going along power line sections and join them as member of a relation.

Circuits are first of all linear between two points and can also be branched as to involve more than two substation in them, always limited by circuit breakers.

Power line sections

A power line section is defined by 601-02-30 IEC60050 definition. It consist in one or more continuous line segments (a consistent part of a power mapped as a way with tags in OSM) forming a link between substations or tap points. A section always have exactly two ends.

In practice, we should find continuous conductors going along power line segments between substations or tapping points.
For sake of mapping simplicity, describing a section as a relation will only be required when this section if composed of several segments that need to be combined.
Otherwise, the section and segment are merged into the existing power=line way without further change.

Power circuits aren't routes

Power circuits are not similar to transportation routes. Power lines allow power to flow over them and aren't similar to transportation networks with intermediate stops.
Also there are no traveling things like buses, passengers or travelers as in "common" routes. A traveling of electrons is rather physical abstraction, and it is not so simple as "small moving balls".

Proposed tagging will enable a better split between transportation and power software as the last will only rely on power=* and won't have to look for any route=* in input datasets.

So the structure of this relation is very different from structure of a route relation.
That is why we should not tag power circuits with type=route + route=power. We should think about a special relation, and I suggest type=power + power=circuit for this purpose.

There is no trunk and branch

This section relates on circuits with more than two ends.
Some models sometimes define a multi-tenant circuit as a trunk line and one or several branch lines. Such a distinction isn't relevant as the trunk line may be any valid combination of line sections linking two substations through the given circuit. Nothing but the geometry of lines allows to set the trunk line as the one which goes on without change in its direction (if applicable).

We'd better defining only line sections which converge on tapping points and join them in a proper circuit relation.

If and only if a given section is composed of several segments, an intermediate relation could be used to summarize the physical properties (defined below) of the whole section instead of qualifying each individual segment.
Otherwise, when a section is composed of a single segment, this segment () can be directly involved as a member of the circuit relation with role section.

Physical properties

All power lines aren't equivalent when it comes to make power flows along them.
They are first of all designed for a given voltage=* we are used to document in OSM.

Power lines have two important physical properties: resistance and reactance. Together, they form the line's impedance, which is a quantity that describes how the line resists the flow of alternating current.
Resistance accounts for energy losses due to heat, while reactance arises from the line's inductance and capacitance.

How will we get such information?

Resistance and reactance values depends on the nature of conductors, mainly their material and cross section.
At first, they are barely guessable from ground unless conductors nature would be printed on poles or wherever else. That's why refining power=line with material and cross section tagging will lead to poor results.

Public information, scientific datasets or open data could help to find them resistance and reactance as many scientists and researchers currently build static network models.
They should be handled carefully, particularly about licensing and accuracy questions.

Proposed keys are obviously optional and a placeholder to organize consistent knowledge.
We choose to add them on the proposal as to define explicitly how they combine when define both on power line sections and circuits relations.

Why aren't we prompted for capacity in MW?

Public communication often deal with actual power line capacity in MW as the amount of power that could flow in the power line without damage.

Such values aren't constant and varies along seasons and upon operational conditions. They're sometimes restricted in a given time period (i.e 4 000 MW during 10 minutes) as to not overheat and finally destroy the power conductors.
We shouldn't add this value to OSM despite interesting as it won't be accurate.

Less tagging redundancy between lines, sections and circuits

This proposal is an opportunity to solve some redundancy issues we had between lines and circuits from the beginning. Just like other fields of knowledge, using relations on top of a physical network of lines allows to get a more precise meaning of tags.
See the following table summarizing how tagging should be:

Tags	Meaning on lines	Meaning on sections	Meaning on circuits	Comments
frequency=*	-	-	The frequency at which the circuit is operated	Frequency is an operational value independent from physical conductors. It only matter for circuits.
wires=*	Conductors bundles arrangement	-	-	Bundles arrangement only regard physical line. A given circuit can go through several line sections with different bundles.
cables=*	How many cables on the physical line	How many cables the section involves	How many cables the circuit involves	Cables can be used on both. The sum of all operating circuits over a given line should be <= of the line total amount of cables.
voltage=*	The maximum voltage the line has been designed for	The maximum voltage the section has been designed for	The operating voltage of the circuit	Circuits can be operated at lower voltages than permitted by the actual line design
name=*	-	The line section's name	The circuit's name, usually involving substations' names at its ends	Circuits will combine several line sections and they should all be named after what they actually represent
resistance=*	-	The line section's own resistance	The circuit's global resistance (if linear only)
reactance=*	-	The line section's own reactance	The circuit's global reactance (if linear only)

Tagging

Circuits relations

Key	Value	Comment	Recommendation
type	power	This is a power relation	Mandatory
power	circuit	This relation represents a power circuit	Mandatory
voltage	<Operating voltage>	The voltage at which the circuits operates, in volts	Recommended
frequency	<Operating frequency>	The frequency at which the circuits operates, in Hertz	Recommended
cables	<Circuit's cables>	The amount of cables involved by the circuit	Recommended
resistance	<Circuit's resistance>	If linear, the circuit's global resistance in Ohm	Optional
reactance	<Circuit's reactance>	If linear, the circuit's global reactance in Ohm	Optional
ref	<Reference>	The circuit's reference	Optional
name	<Name>	The circuit's readable name	Optional
operator	<Company name>	The company in charge of circuits operation	Optional

Relation's roles

The relation combines one or more several line sections and substations in which the circuits originates and ends

Member's role	Member's type	Count	Member	Description
section		one or more	Power line section	A section of the power circuit
substation		two or more	Substation	A substation in where the circuit starts / ends

Change management

Affected pages

• Edit type=* page
• Edit power=circuit page
• Create power=line_section page
• Create type=power page
• Create resistance=* page
• Create reactance=* page

Tags to be replaced

Obsolete tag	Usage	Used for ?	New tag(s) to use
type=route + route=power	23 484 on 2025-03-21	A route used to describe a power circuit	type=power + power=circuit
frequency=*	45 654 ways on 2025-03-24 query	Frequency mention on power line sections members of an existing power circuit relation	Remove frequency=* from the line section
wires=*	12 447 relations on 2025-03-24 query	Wires mention on existing power circuit relations	Remove wires=* from the relation

External discussions

• See all previous comments in power routing proposal

Examples

A simple circuit with two ends

This circuit is said simple because it links two substations with no branch. However, it runs over 5 different line sections and you will observe that the wires changes along its path.
We see such topology because of successive power grid improvement plans.
We are able to complete physical properties of the circuit only from JAO static grid model published online.

Key	Value	Comment
type	power	This is a power relation
power	circuit	This relation represents a power circuit
voltage	400000	The voltage at which the circuits operates, in volts
frequency	50	Alternative public grids are operated at 50Hz in western Europe
cables	3	It's 3-phase without neutral power circuit
resistance	1.65	The circuit's global resistance in Ohm from JAO dataset
reactance	21.04	The circuit's global reactance in Ohm from JAO dataset
ref:FR:RTE	CORNIL71M.LAN	The circuit's reference
ref:EU:ENTSOE_EIC	17T-FR-00000066P	The circuit's reference
name	Cornier-Montagny les Lanches 1	The circuit's readable name
operator	RTE	French transmission grid is operated by RTE in France

Sections A to F are power=line ways and will get the section role. Substations are power=substation members with the role substation.
This circuit is currently described in 5459750 5459750 that will need to be refined, because using the discouraged type=route tagging, if this proposal gets adopted.

The difficulty to produce an accurate model of the actual power path with the physical lines knowledge only appears clearly.

Voting

Voting has not started yet