Proposal:Power circuits routing
| Power circuits routing | |
|---|---|
| Proposal status: | Draft (under way) |
| Proposed by: | fanfouer |
| Tagging: | power=circuit |
| 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 what actual paths power can follow 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 and without majority consent. This proposal represents an effort to push forward key elements and enhancements that could be achieved at the moment.
This presentation made on SOTM-EU 2023 intends to bring a summary of what it is proposed and what matters most.
This proposal is part of the work program of the Oh my grid initiative.
Proposal
This proposal seeks to approve two new relations type=power + power=circuit and type=power + power=line_section.
These relations involve elements such as power lines (in a broad sense, including overhead lines, cables, and minor distribution lines) and substations that form a continuous electrical circuit—i.e., a physical path through which electricity flows across the grid.
The latter relation type=power + power=line_section will allow mappers to define a subpart of a power circuits (a section) which may contain one or more power line segments as shown in the "Rationale" section below. This provides a more flexible and semantically meaningful structure than relying solely on linear features or creating single-member relations for short segments.
To complement the above, it is also proposed topology=*, to indicate the structural layout of the circuit. This would distinguish between:
- topology=linear: circuits with two endpoints (a typical point-to-point transmission line),
- topology=branched: circuits with three or more endpoints, such as ring or mesh networks.
Finally, to enable power systems analysis and electrical modeling enhancements using OSM data, it is proposed to add (optionally) three physical parameters applicable to both power lines and circuits:
- impedance=*: expresses the total impedance of the circuit or line, quantifying the opposition to alternating current in ohms (Ω).
- resistance=*: represents the conductor’s opposition to direct current (in ohms).
- reactance=*: defines the inductive or capacitive reactance, i.e., the opposition to alternating current due to inductance or capacitance (also in ohms).
By including these three physical parameters the community will be supporting a more accurate power grid modeling, facilitating integration with simulation tools, and improving semantic clarity for infrastructure mapping in OpenStreetMap.
Rationale
As OpenStreetMap represents physical 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 roads ways.
Mapping power circuits as relations is an established practice since 2017. The point of this proposal is to promote a single tagging model instead of several (at least two, maybe more).
It is important to express that this proposal does not aim to increase unnecessary complexity or introduce excessive use of relations. Its authors and many in the power infrastructure mapping community are committed to finding the cleanest and most maintainable solution possible. Nevertheless, in certain cases, the use relations is required to provide a robust description of logical concepts over physical assets of power grids without information redundancy.
Power circuits
A power circuit is defined by 601-02-28 IEC60050 definition. It represents the longest continuous portion of a power grid between circuit breakers, which are typically located within substations. Circuit breakers isolate electrical faults occurring on a given circuit to avoid the propagation of the fault to other circuits and prevent power grids failure.
A circuit is composed of as many power line sections as necessary.
In practice, a circuit can be mapped by finding continuous conductors, going along power line sections without breakers, and joined as members of a relation.
Typically, circuits are primarily linear, connecting two endpoints. However, they can also be branched, linking more than two substations, as long as the continuity is maintained.
Power line sections
A power line section is defined by 601-02-30 IEC60050 definition. It consists of one or more continuous line segments (i.e., consistently tagged power=line ways in OSM) that form a connection between substations or tap points.
Each section always has exactly two endpoints. In practice, a line section is composed of conductors that run continuously along power line segments between substations or tap points.
To keep mapping simple, a relation is only required when the section includes two or more segments that need to be grouped. If the section is composed of a single segment, the concepts of “section” and “segment” are merged, and no relation is necessary—the single power=line way is sufficient.
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 power circuit relations 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 tag power circuits as a dedicated relation type=power + power=circuit as it is suggested in the proposal.
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 be 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 other important physical properties: resistance (IEC 131-12-04) and reactance (IEC 131-12-46). Together, they form the line's impedance (IEC 131-12-43), 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.
Theses values are expected in Ohms for the whole segment or section, not in Ohm per kilometer or any distance unit.
How will we get such information?
Resistance and reactance values depend on the nature of conductors, mainly their material and cross section. Impedance is a combination of both reactance and resistance.
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 find impedance, resistance and reactance as many scientists and researchers currently build static network models. The found data should be handled carefully, particularly about licensing and accuracy questions.
Proposed keys are obviously optional and a placeholder to organize consistent knowledge. We have chosen to add them on the proposal as to state explicitly how they combine when defined both on power line sections and circuits relations.
How will we update such information?
Physical properties of power lines are subject to occasional change, particularly as the capacity of power lines can be improved by reconductoring or rewiring. In simple words, reconductoring a power line is the replacement of a traditional/old transmission line with new conductors. As a result, system operators often practice reconductoring to achieve more resilience or power flow capacity in their network.
To update resistance, reactance and impedance, open data, specialized media or operator's communication could be used to help us detect such changes and properly update OSM.
It is possible to automatize the conflation of some sources with QA software to prompt for mappers to update depending on what exists in OSM.
As upgrading require heavy works along power lines, some mappers may notice such a rewiring and at least add a note allowing anyone knowledgeable to look for up to date information.
How to aggregate properties?
Physical properties of line, such as its length, can be summed over continuous segments but won't be aggregated on branched circuits.
Branched circuits won't get physical properties and values will be given on each linear section involved.
For instance, if topology=linear and line segments are connected one after the other, reactance, resistance or impedance could be summed directly.
However, if tow or more circuits or two or more sections exist between two nodes/substations, grid topology and network analysis tools are needed to calculate overall impedance. This falls out of the scope of the proposal which is why only linear sections will be given physical properties (optional) when found.
Why aren't we prompted for capacity in MW?
Public communication often deals 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 vary due to 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 and lead to wrong interpretation.
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 | Single segment merged with section | Meaning on circuits | Comments | |
|---|---|---|---|---|
| Meaning on line segments | Meaning on sections | |||
| 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 |
| impedance=* | Recommended for sections only | The line section's own impedance | The circuit's global impedance (if linear only) | |
| resistance=* | Recommended for sections only | The line section's own resistance | The circuit's global resistance (if linear only) | |
| reactance=* | Recommended for sections only | The line section's own reactance | The circuit's global reactance (if linear only) | |
Tagging
Power lines
Power line segments mapping isn't changed by this proposal and practices explained in power=line, power=minor_line and power=cable remain valid.
Pages will be updated to explain it's about mapping line segments, as per 601-02-31 IEC60050 definition.
Proposed keys impedance=*, resistance=* and reactance=* will be added to available tags.
Circuits relations
| Key | Value | Comment | Recommendation |
|---|---|---|---|
| type | power | This is a power relation | Mandatory |
| power | circuit | This relation represents a power circuit | Mandatory |
| topology | linear or branched | Nature of power circuit topology | Recommended |
| 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 |
| impedance | <Circuit's impedance> | If linear, the circuit's global impedance in Ohm | Optional |
| 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 |
Defining physical properties on branched circuits (with topology=branched and/or with 3 or more substations as members) isn't allowed and mappers should be discouraged to do so.
Relation's roles
The relation combines one or more line sections, 0 or more tap points and two or more 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 |
| tap | 
|
one or more | Power line tap, related to line_management=branch | A node at which several sections connect to form a tap point |
| substation | 
|
two or more | Substation | A substation in where the circuit starts / ends |
Circuits topology
The key topology=* is intended to be generaly used on any relation on which the topology nature need to be explicitly stated.
| Key | Value | Comment |
|---|---|---|
| topology | linear | The relation involves several sections in a given order and continuous manner, with exactly two ends |
| branched | The relation involves several sections that aren't continuous with 3 or more ends |
Sections relations
| Key | Value | Comment | Recommendation |
|---|---|---|---|
| type | power | This is a power relation | Mandatory |
| power | line_section | This relation represents a section of a power line | Mandatory |
| voltage | <Operating voltage> | The voltage at which the section operates, in volts | Recommended |
| cables | <Section's cables> | The amount of cables involved by the section | Recommended |
| impedance | <Section's impedance> | The section's global impedance in Ohm | Optional |
| resistance | <Section's resistance> | The section's global resistance in Ohm | Optional |
| reactance | <Section's reactance> | The section's global reactance in Ohm | Optional |
| ref | <Reference> | The section's reference | Optional |
| name | <Name> | The section's readable name | Optional |
| operator | <Company name> | The company in charge of section operation | Optional |
Relation's roles
The relation combines two or more line segments.
A relation with a single line segment is not permitted and mappers should be discouraged to add them. They will prefer involving the power=line segment directly in a circuit relation.
| Member's role | Member's type | Count | Member | Description |
|---|---|---|---|---|
| line |
|
two or more | Power line segment | A segment of power line involved in the power section |
Change management
Affected pages
- Edit power=line page to add segment definition and physical properties
- Edit power=minor_line page to add segment definition and physical properties
- Edit power=cable page to add segment definition and physical properties
- Edit type=* page
- Edit power=circuit page
- Create power=line_section page
- Create type=power page
- Create topology=* page
- Create impedance=* 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 |
| power=branch | 244 on 2025-03-26 | A relation to link trunk and branches | To replace by power=circuit and case by case required power=line_section relations |
| 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 linear circuit with two ends
This circuit is said simple because it links two substations with no tap point.
However, it runs over 5 different line sections and you will observe that the wires changes along its path.
As the circuit only have two ends, it is possible to directly involve line segments in the circuit relation that will hold physical properties of the single section.
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 |
| topology | linear | This power circuit is linear with 2 ends |
| 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.
A more complex circuit with sections as single segment
This circuit involves 3 line sections and 1 tap point to link 3 substations together.
All 3 sections are composed of 1 power=line segment, so no additional power=line_section is required.
It is still possible to involve the ways as circuit relation members and section role.
It is not relevant to define physical properties on branched circuits as they couldn't be merged linearly.
Each power section will get its own figures coming from the static grid model.
| Key | Value | Comment |
|---|---|---|
| type | power | This is a power relation |
| power | circuit | This relation represents a power circuit |
| topology | branched | This power circuit has 3 ends so form a branched topology. 3 substation members are expected |
| voltage | 225000 | 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 |
| name | Grandval-Lanau-Rueyres 1 | The circuit's readable name |
| operator | RTE | French transmission grid is operated by RTE in France |
Sections A to C are power=line ways and will get the section role. Substations are power=substation members with the role substation and tower #1 will be member with role tap.
This circuit is currently described in 5465785 5465785 that will need to be refined, because using the discouraged type=route tagging, if this proposal gets adopted.
The most complex circuit with sections as relations
This circuit involves 3 line sections and 1 tap point to link 3 substations together.
The northern section is mapped as a relation as it involves 3 different segments, with different tags that can't be merged into a single one.
Two remaining sections are involved in the circuit relation as ways and get the section role.
It is not relevant to define physical properties on branched circuits as they couldn't be merged linearly.
Each power section will get its own figures coming from the static grid model.
Here is the circuit relation:
| Key | Value | Comment |
|---|---|---|
| type | power | This is a power relation |
| power | circuit | This relation represents a power circuit |
| topology | branched | This power circuit has 3 ends so form a branched topology. 3 substation members are expected |
| voltage | 225000 | 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 |
| name | Saint Auban-Salignac-Sisteron 1 | The circuit's readable name |
| operator | RTE | French transmission grid is operated by RTE in France |
Here is the section C relation:
| Key | Value | Comment |
|---|---|---|
| type | power | This is a power relation |
| power | line_section | This relation represents a power line section |
| voltage | 225000 | The voltage at which the section operates, in volts |
| cables | 3 | It's 3-phase without neutral power section |
| operator | RTE | French transmission grid is operated by RTE in France |
| ref:FR:RTE | SISTEL61ZSAL6 | The local identifier of the power line section |
Sections A and B are power=line ways and will get the section role. Section C is a power=line_section relation and will get the section role.<
Substations are power=substation members with the role substation and tower #1 will be member with role tap.
This circuit is currently described in 5977379 5977379 that will need to be refined, because using the discouraged type=route tagging, if this proposal gets adopted.
Voting
Voting has not started yet