Organised Editing/Activities/Glint Solar power mapping

From OpenStreetMap Wiki
Jump to navigation Jump to search

Rationale

We aim to contribute to power line and substation mapping in the United States.

Glint Solar provides solar developers software to help them build their projects as soon as possible. One of the most important aspects of site selection for solar parks is distance to electrical infrastructure. The better the maps are, the easier it is for stakeholders (including our customers) to make good decisions. Hence, we would like to help improve OSM.

The high voltage lines are already mapped quite well. Most of our edits will be adding distribution lines and substations. Hence, our most used tags will probably be power=minor_line and power=substation. Industrial scale scale solar parks aren't typically built within settlements, so we will have our focus on mapping areas outside of settlements.

We decided to focus on the US due to business considerations.

Contact

zabop (message on osm, edits, forum, email, changeset comments)

Community consultation

Announcement on community forum.

Hashtag

#glintsolar

Timeframe

16 Dec 2024 onwards. We don't yet have a set end date.

Tools and data sources

Editors: JOSM, iD, Level0.

Data sources: default imagery sources in JOSM and iD.

Participants

This list will be extended when others confirm their participation.

Measuring our success

OpenInfraMap's Stats page shows the total length of power lines in the United States. We want this number to increase. This page is only updated once in a while. To measure our impact on a shorter timescale, we prepared these tools:

Of course, none of these tools provide numbers the maximization of which would serve as our goals. They are just metrics on how we perform. The goal is to improve coverage of mapped power lines.

Some issues to keep in mind when using these tools: splitting a power line in 2 typically introduces a new way, and the Changeset Inspector site measures its length. No more power lines are mapped just by splitting a way, but one of the outputs from the split is still represented as a new way in changeset XMLs. The Power Map State site includes everything touching the relation, so it is possible that by splitting a way outside the relation's boundaries, we have shorter total power line length touching the relation. Again, actual mapped power line length has not changed.

The two tools above seem to be consistent. Take Changeset 159770295 as an example. Changeset Inspector says: "Total length of new ways: 38804 metres". Given that it's fully in Arkansas, we can use Arkansas' relation ID 161646 and Power Map State site to check total length of power lines in Arkansas before and after the changeset (I use 29/11/2024, 00:00 and 01/12/2024, 00:00). Total power line lenght before: 15731.670km; after: 15770.477km. The difference is ~38.8 km, a value very close to the output of the Changeset Inspector.

Pull Requests to these tools or suggestions for alternatives are very welcome.

Training/instructions

Our current plan for mapping strategy is: 1) find a place where we expect to have power lines but these aren't fully mapped. These can be substations and other places where grid lines end. OpenInfraMap is an ideal starting point. 2) Try find poles or their shadow, and follow the gridlines. Try staying outside of towns and villages.

Once we have scheduled meetings, we will post access details here, so anyone can come and join them.

Post-event clean-up

We plan to work with experienced mappers primarily. When this is not the case, we will keep an eye on their changesets. We listen to feedback from the users of our software, and if they note something is suspicious, we are going to check it, even if the feature concerned is not the result of one of our edits. We are not planning on having a fixed end date to our contributions - it's our continued business interest to keep quality of OSM power data as high as possible.

Results

We will share details about our progress here. We plan to make charts showing how total lengths of power lines changed in an area over time. If those progress reports are overly superfluous to be included here, we will link to them from here. First report expected on 13 Jan 2025.