October 03, 2023
By Michael Barnard

Midjourney generated image of world laced with transmission wires and sparks

Wind turbine repowering project comes with agreements for landowners

There are about 200 GW of HVDC cables stretching 58,000 kilometers in operation today, and about 180 GW planned stretching about 45,000 kilometers in planning and construction.

One of the big levers for decarbonizing our global energy is by connecting continent-scale grids with high-voltage direct current transmission (HVDC). We’ve been doing a lot of that already, but a lot more is planned. There are about 200 GW of HVDC cables stretching 58,000 kilometers in operation today, and about 180 GW planned stretching about 45,000 kilometers in planning and construction. The large majority of what’s in operation is in China, of course.

Let’s start with the why. As delayers love to say, the wind doesn’t blow all the time and the sun doesn’t shine all the time. But go north, south, east, and west a few hundred or a couple of thousand kilometers, and the weather is completely different. We’ll be overbuilding renewables in the best places and connecting them to the highest demand centers like major cities, ports and industrial sites. The sun may not shine brightly during peak demand periods of the late afternoon and early evening, but it’s shining brightly a couple of time zones or more to the west.

But traditional transmission using high-voltage alternating current (HVAC) loses a lot of energy over longer distances, and a lot more if you have to go underwater or underground with the cables. HVDC loses a lot less energy over great distance, underwater, and underground, about 3% for every 1,000 kilometers. That’s the most efficient mechanism for transmitting energy we’ve discovered.

It’s so effective and so efficient that it’s hard to keep up with all of the projects. Enter RTE International, a consultancy and engineering company whose activities cover all areas of electricity transmission. They keep track, as best as they are able, of all existing and planned HVDC projects globally. They publish updates as part of their monthly HVDC newsletter, something I’ve been receiving and digging through for a while now.

A big part of the newsletter are the huge tables of all existing and planned HVDC projects. I’ve been looking at it for a while and wishing I could do some basic analysis on it. After a quick conversation with the people at RTE responsible, I pushed it through a PDF to spreadsheet converter, cleaned up the data, made some guesses and started doing some cross tabs.

I’ve chosen to simplify some things in the process, and make some choices about what the data signifies. That means that some things have become less accurate. For example, the RTE data set includes the lengths of a specific project that are underground, subsea, or on overhead lines (OHL), and many projects combine all three for different portions of the route. I’ve simplified that a bit, so any cable with a subsea segment is represented as all subsea for example. It’s imperfect, but I’m interested in gross numbers.

Similarly, the RTE data has the number of lines of all different sizes, and I’ve chosen to separate that into total power and the number of lines, making the assumption that future projects with only the power have only one line. That’s inaccurate I’m sure, but the errors are understood. Similarly, where no power in MW was expressed or distance in kilometers was listed, I put in guesses. Finally, where future dates weren’t listed, I chose 2030 to have some data in the mix.

All of that data ‘cleansing’ means that any errors are most likely mine. That said, RTE’s data isn’t perfect. I’m aware of numerous existing and planned HVDC interconnectors in the ASEAN region, most to and from Singapore, but they are missing from the data at present. There is an emerging ASEAN supergrid that I’m paying some attention to which appears to be off of RTE’s radar. And as I found out while working on this analysis and writing as I went, it appears to be missing a lot of data from China.

Locations of current, planned and in construction HVDC projects globally from Open Street Maps

Wikipedia maintains a list of the HVDC operational projects and some planned projects, and someone kindly pushed the data into Open Street Maps to create this visualization. It leaves some questions open. For today’s effort, I mostly didn’t try to integrate Wikipedia’s data with RTE’s and resolve the overlap.

But RTE’s data did allow me to ask a few questions I’d been wondering about. The current and future GW and kilometers listed in the opening paragraph, combined with China’s data set from Wikipedia, was one of them.

Another question was the breadth question. Remember that the intent of transmission is to bring electricity from where it happens to be generated to where it is needed. As such, future projects would reasonably be expected to have both a longer transmission length and be more likely to connect countries.

And that’s true. Even with my choice to use 100 kilometers as the length for projects without any length data (all future), the average length of future projects is about double historical projects — 454 kilometers vs 229 kilometers. (The lack of a lot of China’s operational data in RTE’s data cuts those numbers down a lot.) HVDC projects of the future are running much further than historical projects, at least outside of China. Somewhat surprisingly, current projects aren’t necessarily connecting countries more than historical projects as a ratio. About a quarter by power of the projects was historically flowing between countries, and that’s true for future projects as well.

Of course, there is a reason for this. The majority of all operational projects and a fifth of future HVDC in the world is in China, about 170 GW existing and 27 GW planned or under construction. China’s a big country, the third largest in the world by land mass and slightly larger than number four on the list, the United States. The USA has about a GW of HVDC today, about 2% of the global total, and a 35 GW planned per RTE.

I cross-checked against the Wikipedia data set, and found a lot more Chinese projects than the RTE data has. Wikipedia has 55 HVDC lines in the country, while RTE only shows 18 between operational and future projects. All 55 are operational per Wikipedia, which has different imperfections than RTE’s data, of course. But assuming that Wikipedia data is more correct for operational Chinese HVDC, and RTE’s data is more accurate for the rest of the world and future projects, the country has more than the rest of the world combined, which was my assumption going into this analysis exercise.

Assuming RTE is more right for the rest of the world than it is for China (a fairly safe assumption), then the total HVDC globally is about 200 GW totally, with 72% of that inside China, and the total current length of HVDC is about 58,000 kilometers, with 86% of that inside China.

It’s taken me a long time to find analysts who read Mandarin and follow China’s energy progression from the inside by reading the country’s reports in their own language, and I’ve had conversations with international experts like Bent Flyvbjerg about the challenges. No flies on RTE if their China data is less complete. And, of course, the USA is the country of aspirations, so I suspect less of that future HVDC will be completed as an average. But it would make sense for RTE’s people to cross-check with Wikipedia and reduce the disparity given that they are pushing their HVDC data set out monthly.

The point of that is that China and the USA having a third of the planned HVDC power between them means that continent-scale grids are being built, just inside countries that are the size of continents. China’s average operational length is about 1,300 kilometers per Wikipedia, while its future length of cables is about 1,400 per RTE. The variance is even more extreme for the USA, with an average of operational cables of 63 kilometers vs a future project average of 420 kilometers (all RTE numbers).

There were some other things I expected the data to show. HVDC has been historically used to link grids together. Operational grids have a rhythm, the beats per second that the current alternates, typically 50 or 60 times a second. HVDC has no rhythm as it doesn’t alternate. That makes it a great tool to connect two grids so that they can share electricity without having to synchronize their beats per second which are either simply misaligned by a fraction of a second, or have a different beats per second entirely. That’s referred to as back-to-back HVDC connections and it’s usually done at the geographical point where the grids meet, so the distance of the linkage is zero.

I expected that operational HVDC systems would have a higher percentage of back-to-back HVDC connections than future grids, and RTE data does show this. Historically, about 18% of HVDC operational systems by power were back-to-back, but in the future only about 3% of planned systems are back-to-back. As a note, in the past vs future length assessment above, I removed the zero distance back-to-back connections.

I also expected more future projects to be subsea, and even with China’s massive in-country projects excluded, about 5% more of future projects are subsea. That’s another indication of connecting countries as opposed to being inside of them.

HVDC being the new pipeline continues to be supported by this. As we look around the world, HVDC projects are booming everywhere. The aluminum that they are made of is the most common element in the earth’s crust, and new HVDC factories are being built on multiple continents. They are moving TWh of renewable electricity across thousands of kilometers today, and will be carrying vastly more in the future. Countries which continue to focus on moving molecules for energy are going to be left behind.