# Why are frequency= and voltage= a tag of OSM ?

Dear OSM fellow contributors,

In my understanding, OSM is supposed to collect data about the physical world, not the electrical world. In an overhead-line, or in a high-voltage power line, the frequency of the current is not part of the infrastructure. We *could connect the conductors to a power source of any frequency, and it’ll work (). Nevertheless, there’s a frequency tag in OSM.

Voltage and current are more tricky. They are not themselves part of the infrastructure, even though it is true a fixed voltage and variable current is often used. For example when the simply shut off a power line for maintenance, the voltage=0 and as such the data in OSM becomes “wrong” even though the infrastructure hasn’t changed !

The infrastructure allows for a maximum voltage and maximum current. If the maximum voltage is exceed, it will become dangerous and electrical arcs could appear. If maximum current is exceeded, the conductors could overheat. So maximum voltage and maximum current are part of the infrastructure (mappable of OSM), but voltage and current are electrical properties that can vary and should not be mappable on OSM.

Some high-voltages lines where I live use a mixing of different voltages frequencies. For instance this one. Some conductors in the line are used to power trains (16.67 Hz) and some to power civilians (50 Hz). As such the taging is confusing. Even the “maximum” voltage, which is part of the infrastructure, would be different depending on the conductors.

I’m sorry if it sounds like I’m complaining about tiny details.

(*) Ok this is false, a too high frequency will very quickly create RF emissions and related problems, but my point is that the wires doesn’t by themselves a frequency - yet a frequency is tagged on OSM.

Yes, we do tag frequency and voltage:

Regards
walter

I know but that wasn’t my point. My point is that frequency and voltages are not observable on the ground, and as such, should not be part of OSM database according to OSM “on the ground” philosophy. Or you could just answer me that it’s an exception and that I’m a moron - and I already know that.

In practice though a given power line (or other component) will be used to carry a fixed nominal voltage and frequency, with the current varying according to demand. Just like maxspeed on roads will vary according to the conditions (traffic, roadworks, weather etc).

The example you quote, is what it is. Actually it would be two sets of wires, sharing pylons I guess, and never connecting in any way. How would you suggest mapping this otherwise?

Maxspeed is very observable from the field, also it is tagged “maxspeed=” and not “speed=” which is correct. OSM seems to be mapping “voltage=” for what is either maximum or nominal voltage, but it’s definitely not always what you’d measure if you had a giant voltmeter and measured the voltage here. Doing so is similar to if a road would be tagged “speed=” instead of “maxspeed=”.

It was just an example of shortcomings. The wires just carry current from point A to point B and don’t have intrinsic voltages nor frequency. They could decide to modify the connexions at substations and uses wires used for trains for civilians or the other way arround, and the frequency would change, but the infrastructure would be the same. Similairly, for whatever reason they could decide to change the voltage anytime, including to shut off the line (that is, voltage=0).

Similarly, the overhead catenary of a train line could change voltage or frequency to accommodate for a different rolling material - but the infrastructure would stay the same.

If I had to suggest something, voltage= and frequency= tags should be used on substations only, not on power lines themselves which does not have any intrinsic voltage or frequency.

That will not become easier when maximum values are demanded.

Although the idea is in itself not strange you’re actually asking people to remove all frequency and voltage tags because nobody knows their maximum values.

1. Occasionally, power lines are not powered for one reason or another for periods of time. Should the lines’ tags be changed to frequency=undefined and voltage=0 during those periods. I, for one, would neither expect nor desire that such changes be made.

2. The presence of power, frequency of alternating current and good estimates of the operating voltage can all be done “from the ground” by measuring the electric and magnetic field characteristics. Power companies often have such equipment. Such equipment are quite legal for private individuals to own and some do. Distance from the ground and sizes of insulators are clues to the operating voltages. The effective diameter (see skin effect; there is a discussion of this issue for power lines in the Wikipedia article) of the power lines provides clues to the operating frequency–the higher the frequency; the larger the diameter. See Hall effect for measurement of magnetic fields. See electrical field measurement techniques. Commercial electrical field meters do exist.

3. Lay people can learn easily to identify visually the general (e. g., the power of 10) designed or nominal operating voltages and currents of power lines. This knowledge then can be used for identifying such a feature on a map, if present, for orientation purposes.

4. I would not be surprised that OSM data included the locations of subterranean features such as subways, vehicle and pedestrian tunnels, water mains and sewers, power and communication cables, etc. I would not be surprised that OSM data included the locations of near surface features and airways as well.

5. OSM allows, permits and encourages users to select which data is desired. If you do not desire or see a use for certain data, then do not enter it yourself (perhaps, someone else will enter it later and you should not then remove it and do not extract it from the OSM database. Remember that someone else may be using data for which you see no use. Only remove data that is wrong.

Argh it seems no matter I say people misunderstand me 100%. I do not speak good english so please be nice. I never had the intention to remove the data, or to alter it temporary when a line is shut down. I wanted to point out that his data shouldn’t have been part of OSM and discuss the concept of why those tags exist in the 1st place.

But in all cases this is pointless and I should just accept things as they are, no matter whether this makes sense or not - and I regret making this thread in the 1st place.

1. I do not believe that I completely misunderstood you.

2. Your largest error is contained in “…that [t]his data shouldn’t have been part of OSM…” That is wrong. That is why you received such as strong response.

3. This discussion is not pointless. People tried to inform you of some of the reasons that those data were there…

4. Yes, "I[You] should just accept things as they are, no matter whether this makes sense or not to you.

5. I am sorry that you regret making this thread in the first place. I hope that you learned sometime useful from this thread.

6. English not my first language. I lost my first language as a very young child when my great-grandmother died. That language was German.

7. I spent the last quarter century working in a heavily multi-lingual environment. Eighty-seven languages were known to be used on site. Your English is fine. I understood you to mean that you wanted an explanation of why those data were in the OSM database.

8. Power transmission facilities are part of the physical world and many of their characteristics are determined by the voltage, amperage, frequency and phasing of the power they transmit. I have power lines in the maps I make for my older car GPS receivers and at times, I can place where I am at by seeing a high-voltage transmission facility. I assure you that a 130KV, a 650KV and a 1.5MV facility look quite different.

9. OSM has had problems in the past by new editors whom did not understand why certain data were present and set up facilities to remove those data in mass causing many problems. This is not welcome. I am not saying that you would do this.

That’s not true. Many powerlines have the voltage on a warning sign, and most transformers have a specifications sign fixed to it.

OSM tags many things that are not observable on the ground, like boundaries.

Some people have offered good arguments that frequency and voltage can, in fact, be observed. But more fundamentally, there isn’t any strict rule requiring on-the-ground visibility in the first place.

We do have a “map what’s on the ground” section as part of our “Good practice”, but that only says that on-the-ground observation trumps other sources (imports, imagery, …) if there is a contradiction.

We also have the goal of Verifiability, which excludes subjective features. But it doesn’t necessarily demand verifiability on the ground, which would be a much stricter standard.

Mapping objects that can be observed on the ground is certainly what OSM does best. But at the moment, there is no rule only such objects are allowed in OSM (even though quite a few people might support such a rule).

In the absence of a conflict with any rules, all that’s required for people to add frequency and voltage attributes is that they find them useful or interesting.

I’d rather encourage people to ask questions. When things make no sense, they should absolutely be questioned, and possibly changed. Not unilaterally, of course.

OK I guess I spoke wrongly again. It effectively does not matter whether it is observable on the ground or not, because the “on the ground rule” is applicable to whether there’s ambiguity. Here it’s not the case so I shouldn’t even have mentioned that. Let’s forget I ever mentioned that, ok ?

What I meant however, is that pylons and wires does not have an intrinsic voltage nor frequency. They just carry the voltage and frequency they are powered with. Companies could and did change the voltage and/or frequency without changing the infrastructure. This was definitely done for railway electrification - where the power had been altered but not the infrastructure. The infrastructure could even be built with that in mind from the start.

I do not have any “plan”, “suggestion” or anything in mind when it comes to those tags. I just wanted to mention I found this funny that voltage and frequency are tagged on the wires when they should really be tagged maybe at the substation or transformer that powers the wires. Now I agree there’s not much we could do even if anybody agreed with me in the first place.

I agree that in the case of aerial high-voltages lines, the “voltage” tag could be useful in order to have a global idea of the size of the pylons, that could reveal how important the line is and play a major role in cartography and potential 3D representation of the area for instance. However, that is actually using a “voltage” tag to map what in reality is pylon’s size. A true tag specifying size directly would be better suited - even if I fully understand measuring each pylon’s size is extremely impractical. When it comes to railways’ catenary, I don’t even think the pylon’s size changes that much anyway, and those are more prone to voltage changes as high-voltage supply lines.

I think most of us understand what you’re asking, but I think the problem is that the answer to:

…is basically, “because they are”. At some point, someone thought it would be useful to tag the frequency and voltage that are typically found on particular lines, and nobody objected strongly enough that the data was removed.

I agree that the voltage can change, but isn’t the same true for a maxspeed value? A group of people decided that the current maxspeed on a certain street is X today. Next year it can be something else. The same is true for voltage, I do not see a lot of difference between the two tags. We map the value at this moment and will modify it when it changes.

P.S. I never map the voltage of lines, as I don’t know them, but I do map the voltage of substations, which is always indicated on their door. Of course, I do not check whether the sign values is actually used.

Then we change the tags in OSM. It’s the same as when opening hours of a store change. Something changes in the real world, we change it in OSM.
I also do not see the point in this argument.

It’s not the same. If you’re talking about traditional, maxspeed indicators(*), changing max speed requires changing the infrastructure, an employee should go there and actually change the signals so that a different maxspeed is displayed for drivers.

(*) There exist indicators which can change dynamically/electronically. This is another story.

Voltages and/or frequency can however be switched without changing the infrastructure. There’s actually a train line which does this near Geneva.

In my opinion, this is a better approach than mapping voltage on the line themselves. However I doubt every substation in the world has voltage written on it’s door.

That’s exactly my point - it’s possible to change voltage or frequency without changing the infrastructure - i.e. nothing “changes in the real world”.

I have seen that a lot of power lines have the voltage indicated on them, mostly used as a warning on towers or poles, or in concrete plates in case of an underground line. The same information/warning is indicated on outdoor transformers.
Maybe your ideas to improve the database could be discussed in the context of the Project Power networks, which suggests to discuss them in IRC.

I was talking about implicit speed limits like on motorways and within city limits. They are rarely sign posted (although they might be posted on borders)

And that is mapped as well, like https://www.openstreetmap.org/way/548156141#map=17/51.36340/6.17309

That also happens with speed limits. There are lots of roads that have a different speed limit depending on the hour of day or on the weather conditions.
No changing of infrastructure is needed there.

I still see no argument why setting voltage or frequency would not have a place in OSM.

In this case you’re totally right, France recently brought the global limit down from 90 to 80 km/h outside of villages, and this was done without (significantly) changing the infrastructure.

And @Maarten, you’re totally right too.