Show HN: An interactive map showing live wind farm generation in Great Britain

I'm here to answer for WeatherLayers, AMA! I'm super happy to help @robhawkes with this project.

WeatherLayers consists of two products, that can be used either together or separately:

- WeatherLayers GL is a frontend visualisation library with deck.gl layers which can be integrated with common map libraries. The library can be used either with custom self-hosted data or with WeatherLayers Cloud.

- WeatherLayers Cloud is a cloud service providing pre-processed data for visualization from common public weather data sources (NOAA, Copernicus).

Athough it’s a commercial project, there are ways to discuss a potential discount or even a free usage for a non-commercial project such as Robin’s.

> WeatherLayers Cloud is a cloud service

Finally, a cloud service that has met it's true calling :)

its

Amazing piece of data visualization. I work with deck.gl at my day job, and it's really cool to see how far people can push it.

I have a random question - what layer are you using to draw the country borders on the base map? It really stands out, and helps tie the styling together.

Thanks! All the Deck.gl parts are thanks to the work by Jan and WeatherLayers.

The countries layer is using this dataset from MapTiler [1] which is also the service I'm using for all the base maps and vector tile data. I then used interleaved layers [2] in Deck.gl / MapLibre to place the wind layers behind the country outlines.

[1] https://www.maptiler.com/countries/ [2] https://deck.gl/gallery/mapbox-layer

Glad to see another deck.gl user!

deck.gl focuses on vector data, whereas WeatherLayers GL adds raster capabilities. Actually, WeatherLayers GL is mostly custom WebGL shaders that could be implemented outside of deck.gl as well. The major reason why deck.gl is used is that it serves as an integration layer for popular map libraries. I've written deck.gl-leaflet [1] plugin myself.

[1] https://github.com/zakjan/deck.gl-leaflet

Hi Rob! I have a request from an avid user of your visualisations in the Octopus Energy App. Could you please switch the y axis on real time data from Octopus Home Mini from Wh/min to just instant Watts? kWh/day makes sense on larger time scales but Wh/min is pretty meaningless to most people. Also it seems like it would be more readable as a line graph with datapoints, than a bar chart. A line graph would also allow for better interpolation between missing readings too.

Hey! So the Home Mini visuals in the app aren't something I'm personally involved in but I'll make sure to pass on the suggestion for you.

Thank you!

From this map you might think that the UK is currently depending (96.5%) on fossil fuels. But it's not quite that bad: https://grid.iamkate.com/

The real eye-opener (for me) on the latter display is the emissions per kwh over all time - dropped steadily from 500g in (only!) 2012 to a steady 150g or so recently. Encouraging!

This is a good point – I'm not currently accounting for solar in the aggregate and I'm lumping everything else as "non-renewable" which is true but not ideal as it's not all fossil fuels.

I'm going to look at how to be a little more accurate with this, similar to Kate's great website.

I think hydro and tidal power is renewable right?

Sure is! I'm limited by the way the generation is broken down in the source data [1] as it doesn't make any distinction for tidal generation – though I'd argue that's probably a very small amount in the UK. Hydro is already separated though.

[1] https://bmrs.elexon.co.uk/generation-by-fuel-type

Really illustrates the potential benefits of offshore wind farms. It seems like there's a large overall difference between wind speeds on the land and in the water, especially far off coast.

One of the most interesting findings from building this was to see just how important a distributed wind energy system is. The new offshore mega-projects like Dogger Bank are going to make a huge difference, and more onshore development will help with localised generation as well (avoiding transmission capacity issues, etc).

It's been fascinating to watch the various weather patterns over time to see where the gaps are in the locations of wind farms around the UK. It'll be great once we fill in the big hole in capacity off the south-west coast.

I was surprised to see that Wales is mostly orange:

https://en.wikipedia.org/wiki/Wind_power#/media/File%3AGloba...

The UK has had issues with limited transmission line capacity to Scotland which had led me to think that most of the generation potential was up there.

I live in North Wales so can confirm it's pretty windy here and we have a few onshore wind farms around here too. We've also a few decent-sized wind farms just off the coast like Gwynt y Mor.

The Scotland thing is interesting as there's definitely issues with transmission constraints over the B6 boundary, though there's still benefits to building more wind farms for local consumption and/or storage as more batteries (especially) and pumped hydro capacity is added to the mix. Plus the National Grid have plans to expand capacity through to 2030 but I had to admit that's at the limit of my knowledge.

I was looking into this recently and future developments look exciting. There seems to be plenty of shallow water ripe for development similar to the north sea but there is also some focus on floating platforms too.

My casual observation is that it’s generally windy somewhere around the coast, and when not in one area like the north sea it is windy of the south west or south coast. And vice versa.

Once we build out farms around the four axis of the coast im sure the base load generated in aggregate will rise quite a bit.

> just how important a distributed wind energy system is

Every time I look at anything to do with grid scale energy, you see just how important it is to distribute and diversify everything from sources of energy to geographic location to interconnects.

Capitalism with its capacity utilisation targets, just in time delivery etc is not very well suited for a robust network.

Also no NIMBYs on the North Sea. There's some angry fishermen and ecologists but for the most part "out of sight out of mind".

I'm afraid we killed it… I see just a gray base map and the browser is getting a "504 Gateway timeout" from the API:

    This Serverless Function has timed out.
    Your connection is working correctly.
    Vercel is working correctly.

    504: GATEWAY_TIMEOUT
    Code: FUNCTION_INVOCATION_TIMEOUT

Unfortunately yes. I'm working on a fix but it looks like the API I use for the generation data wasn't able to handle the influx of visitors. Not much I can do right now but I did upgrade my Vercel account to get a longer timeout at least.

Hoping it'll be resolved soon.

It's back now… I'll keep an eye on it.

Thanks!

The Elexon API is down again unfortunately. Lesson learnt that I should be storing this data myself in future to avoid it taking down the entire map.

Getting {"error":"Missing generation data"} now

Yeah, sorry about that! Turns out the Elexon API I use for the generation data doesn't like being on the HN front page. Lesson learnt that I should add a proxy store between their API and the map.

No trouble and thanks for the prompt to go have a look at it again, really cool site!

Just wanted to say, as someone who works in the industry, wind FPNs are sometimes incorrect.

Grid have access to "power available" signals, but this is not public data.

Absolutely! I chose FPN data as it's the best available to me in near realtime but as you say, it's not reliable.

This is partly why I've been collaborating with someone on a prediction feature that tries to guess generation using realtime wind conditions and historic actual generation data at various wind speeds.

I'm hoping to improve things as time goes on and I learn more about the underlying data (and its quirks).

This is cool Robin!

Do you also have access to solar data? Eg, I am a UK homeowner w/ a recent solar installation on my roof, would be curious to see how that looks too.

The EU has this tool for viewing solar radiation data and estimating energy yields based on location & angle: https://re.jrc.ec.europa.eu/pvg_tools/en/tools.html

The interface isn't great, but i find it pretty useful sometimes.

Sheffield solar are your guys. Maybe the best publicly available one.

Thanks! So I don't have solar data yet but I plan to first add aggregated solar using something like Sheffield Solar. Long term I plan to predict solar generation at a project/individual level using weather data and generation models.

There's some previous work in this area by others so I imagine my near future will be spent digging through various papers on the subject.

Please would you give a URL for Sheffield Solar.

I found sheffieldsolar.co.uk but could not find any data within that site.

Here you go: https://www.solar.sheffield.ac.uk/pvlive/

Sorry but isn't 'UK' and 'solar installation' an odd couple? My perception, based upon 40 years of western media consumption is that it's hardly sunny in the UK. I'm ready to be corrected though :)

Solar is pretty popular in the UK. Panels have reduced in cost per watt and energy bills have increased so much that it makes it worthwhile for many homes and businesses.

We do get the occasional sunny day contrary to popular belief, but even on cloudy days it's possible to generate a decent amount of energy.

We also have very long days in summer which can make for a decent amount of kWh/day (although the opposite is true in winter)

Solar isn't an area I'm that familiar with but we have no issues generating plenty of electricity from the sun in the UK.

There's a great site [1] that calculates total solar generation around the UK using a mix of real data and forecasts. We were generating at 5GW around midday and that number will only increase.

[1] https://www.solar.sheffield.ac.uk/pvlive/

Out of date but: https://flatline.org.uk/daystats.html

A ~4kW installation gives me up to ~25kWh on a good day in summer. The big factor is actually the precession of the seasons since I'm at 56 degrees north. Retail electricity prices are now over 30p a kWh, so anything you can do to offset that is very useful.

There is definitely an ROI, especially with recent prices (current tariffs are around 33p kWh)

Then there's the likes of Ripple Energy's solar park co-op offering. The levelised cost of buying in is around 2.5p kWh, with another 2p kWh going on maintenance. The return on the 1st (and active) co-op agreement there via a wind farm is 28p kWh, no reason why the solar park would give different returns than the wind one. Here in the UK grid pricing is more or less the same nationally.

Unfortunately there's much less of a return in Winter months due to higher latitudes, though in Summertime the days are longer than in many other countries.

Annual sunshine is around 1300 hours in Glasgow, 1400 hours in Manchester, 1500 hours in Birmingham, 1700 hours in London and Paris, 2500 hours in Rome and 3500 hours in Cairo.

So it's not so sunny, especially as you go north, but films/TV do exaggerate the problem. "I'm from London" regularly gets a comment about terrible weather, "I'm from Paris" does not.

There's very little solar power in Scotland, as you'd expect: https://electricityproduction.uk/plant/solar/map/

(Figures from https://en.wikipedia.org/wiki/London#Climate and so on.)

The weather in London is not terrible per se, it just changes rapidly, so it could start to rain at anytime during the day.

Very cool. I always used earth.nullschool.net to get a sense for global wind patterns; this uses a similar animation.

That's a classic visualisation and for sure an inspiration with my map! The wind data and animated wind layers are from WeatherLayers, which I highly recommend – great people behind it and very easy to set up.

https://weatherlayers.com/

I admit that earth.nullschool.net and other similar sites have pioneered the visualisation style.

However, their main drawback is that they are standalone, can’t be integrated to another application (besides an iframe).

WeatherLayers is built as a tool for developers, to allow for a seamless integration to other applications and allow overlaying with custom data such as wind power plants in this case, to support end-users with understanding the data in natural context.

Nice! It looks like there's a big storm in the Atlantic right now....

Nice! I worked on a similar project in Ireland - https://www.ux-design-awards.com/winners/2022-1-376-rtv-esb-...

This looks great! Is it publicly accessible anywhere? I couldn't find it.

I'm hoping to expand beyond GB and Ireland would be a great next step if the data is publicly available.

I'v seen charts (someone posted on HN as a comment) where it shows generated electricity by type: solar, hidro, wind, gas, etc and by hour, for many countries. Anyone can bring up the link?

There's a few like that but Electricity Maps is a good one

https://app.electricitymaps.com/map

I’m working on one for the United States! It also has a lot more data about grid operations like wholesale pricing of electricity

https://www.gridstatus.io/

Awesome! Just wanted to say that I love what you're doing with GridStatus – I've been keeping an eye on your work on Twitter.

Cool to see windfarm discussion here. I live in the PNW, love where I live but I have a ton of trees so my solar eligible square footage is pretty modest when juxtaposed to my goals of getting off grid. I recently found out that home turbines are a thing, and I'm hoping with a combination of the two I can achieve my goals. That said, we have no data for this sort of thing, so I'm in the domain of a blind investment. I'd love to hear feedback if folks have experience.

Isn't the problem with a home turbine the low capacity? E.g. a typical backyard turbine might be good for 1000-1500W at peak output, and it's still pretty big and wants to be mounted on a 20 foot pole. The good ones are kind of expensive, too.

I'd wonder if you have a lot of trees if you'd ever get much output from a turbine anyway. At a certain point, if you want legitimate solar or wind power, you probably end up finding a spot to clear some tree cover. Grab a pallet of dirt cheap used panels from SanTan and throw 'em out on the ground.

Personally I probably wouldn't even think about wind unless I lived out in eastern Oregon or Washington on the plains, with big enough property that I could get a halfway decent size turbine and mount it far enough from the house so I couldn't hear it.

Fair assessment. There is space for panels, but I suspect I'll need more of them because the duration of exposure is not what it needs to be optimal. This is the turbine I looked at: https://www.amazon.com/ZHIHUASMTBX-Vertical-Generator-Levita...

My idea was to utilize a combination of solar and this.

Windfarm generation Dashboard in Ireland and Northern Ireland: https://www.smartgriddashboard.com

Thanks for sharing! I'm looking to add more countries to the map so I'll dig into the data behind this one.

ElectrictyMap.org uses a parsing engine to aggregate as many publicly available ISO/TSO/BAs to show live generation mix and cross border energy movement data when possible, and falls back to ML generated estimates otherwise.

https://app.electricitymap.org/map?wind=false&solar=false

https://github.com/electricitymaps/electricitymaps-contrib/b...

Beautiful project you've done, really enjoy the visualizations.

(no association, just a fan)

This is amazing, thanks for the detail! I'll set aside some time to have a proper dig through the Electricity Map sources – looks like it will help me expand my map beyond GB.

Feel free to join their Slack as well, friendly folks if you have any questions.

Suggestion: when I click on a windfarm's name, instead of removing all the other windfarms from the map, it should leave them but highlight the selected windfarm in some way.

Thanks for the suggestion. I'm going to be spending some time soon on the UX so I'll add this to the list to look at.

Does anyone know of a good tool for modeling wind potential?

For example, PVWatts [1] makes it very easy to get solar potential for a site (at hourly granularity). But having trouble finding something similar for wind. Ideally want something that is also validated against data from real world wind farms after they're built.

[1] https://pvwatts.nrel.gov/

Have you seen Global Wind Atlas? https://globalwindatlas.info/en

Looks good thanks! Though it's giving system errors at the moment so I'll try back later.

Interesting to see where wind farms look like they're underperforming; i.e. where wind farms in high wind seem to be producing a lower percentage of their theoretical output than others in low wind areas.

Is that a problem with the data accuracy, or is that reflective of the real-world situation (e.g. misalignment of the turbines against the wind)?

(p.s. thanks for your work at Octopus!)

This could be for a number of reasons, such as:

- Wind speeds are from GFS forecasts and cover quite a wide area - The realtime generation data from the balancing system isn't 100% precise and is averages over a 30-minute period - Some wind farms are simply much more efficient, especially large offshore ones - Individual turbines might be down for maintenance and/or not running efficiently - Localised issues like wake effects and turbulence

I'm not an expert in all this though so take it with a pinch of salt – I'm just very interested in it all and love making visualisations about it.

All that said, one of the intents behind creating the map is to highlight exactly this sort of stuff and to spark a discussion around why it might be.

Also another reason - a generation site may be curtailed, meaning that its output is restricted due to grid capacity limitations, and allowing it to export more energy onto the local power network may exceed safe operating parameters for the lines connecting the generation site to the grid.

Therefore in high wind, you would expect that in areas of dense generation, especially those with limited capacity interconnects to the rest of the grid, curtailment may be seen.

I could see this being the case for regional balancing, but presumably a 1GW generation site has a 1GW-capable link to the grid, otherwise that seems like overbuilt capacity?

(nb of course you might overprovision it to some degree on the presumption that, say, only 70% of your generation is usable at any one time; I more mean the numbers like say 16%.)

What's going on with Seagreen? It's got a forecast of 4MW, and a listed output of 545MW at roughly half capacity, in what seems like pretty light winds.

Looks like it's only just been built? https://www.theguardian.com/business/2023/apr/11/worlds-deep...

This is something I need to look into further but it's currently under construction and being tested. I'm pretty sure they have issued "down for maintenance" notices that I need to take into consideration and supersede any reported generation.

Beautiful work. We're having some pretty odd winds here in NL right now and the North of France is also seeing eastern winds, very odd for this time of year and the resulting cold snap is going to impact crops in a wide region. It might even freeze here tonight, which rarely happens this late in the year.

It's very nice visualisation.

The only comment I have is that concentric circles are not the most intuitive way of showing a percentage of a whole. 50% doesn't 'look' like 50%, and looks similar to 30%.

but I'm not sure how else you would do it.

Very cool stuff! I love how you can see visually the impact they make when it's windy. (I'm sure I've seen something else (none map vis. though) with stuff like grid batteries and gas ones but I have no idea where if that helps?)

How come dogger bank is sat out there where there looks to be plenty of wind, but generating zero?

Good question! Dogger Bank is still under construction but I decided to include it on the map for now (it's a constant work in progress). If you click the little "sparkle" button on the right of the map you can see an estimation of what Dogger Bank would be generating today if it was operational.

It's in the works. https://en.wikipedia.org/wiki/Forewind

(A more informative link would be https://en.wikipedia.org/wiki/Dogger_Bank_Wind_Farm .)

I know wind varies a lot with altitude and there are a lot of available maps of wind such as 2m, 10m, 850mb, 250mb, etc. What altitude wind are you showing on the vis and what are the heights of these wind farms?

That's a great point and something I'm not making clear on the map. I've decided to use wind speeds at 100m on the map but there's an argument to offer data at multiple heights, particularly when looking at shorter onshore wind farms.

With that said, some of the wind turbines already reach 190m in height (eg. Hornsea One) and the mega-sized Haliade-X turbines at the upcoming Dogger Bank wind farm are 260m tall!

As Robin says, visualized wind data are 100m height from surface. That's what we've selected as the most appropriate average for wind farms. GFS contains different heights that we could allow choosing from.

Very exciting project, from a fellow ex-woodmacker. Will get in touch

Thanks fellow ex-WoodMacker! I'll keep an eye open