HVDC transmission line to connect three ISO regions | Hacker News Reader

HVDC transmission line to connect three ISO regions

pv-magazine-usa.com

70 points1970-01-0197 comments

I think HVDC is a game changer for renewables. Any long East-West grid connection is going to improve the total fraction of power that can be generated using renewables because effectively windpower = solar power and having a longer East-West stretch makes the amount of sun hours during a day longer. And any North-South lines will help offset local conditions during the day.

The more the better, especially intercontinental.

I wanted Russia to invest in a transiberia-HVDC line and solar collectors instead of a war. They could even sell excess capacity to the countries on their southern border. I think Reddit panned this idea last time I proposed it because the longest HVDC line was only 2100 miles long.

Yes, that would have been a lot better than war. But it doesn't make you look strong as a dictator...

It does, if done right.

But russia does not have a solar plant industry, but they had lots of war assets sitting idle around, so they used what they had and learned to do.

Russia is one of the worst countries for solar power, perhaps bottom ten. It's also one the better connected ones with old school AC HV lines.

Wind too then. I'm not picky. Hell maybe they find a coal deposit nearby the HVDC line. It wouldn't be the best but burning the coal near the location and shipping the energy else where is cheaper than shipping the coal elsewhere.

Why in the world would they invest into such HVDC project when they already have a huge synchronous grid (unlike the US and Canada, which have 5 grids in total) shared by other countries, especially on the southern border. I think there are even plans for Iran to join this grid. They have absolutely no trouble sending energy across their territory to smooth consumption peaks. Add to that abundance of coal/gas/oil (FYI during winters they use thermal power stations to generate both electricity and heat for homes), robust atomic and hydroelectric generation, and almost zero (in relative terms) solar generation.

They only need relatively short HVDC links to be able to send energy from their hydroelectric installations in Siberia to China and maybe to South Korea and Japan, though with the current geopolitical climate the latter is quite unlikely.

HVDCs are increasingly being used inside grids too. Any grid needs high capacity links (AC or DC) to connect its far corners.

HVDC line is cheaper to build and has fewer losses. But you need conversion stations which are expensive and also have some loss.

At some distance, HVDC trumps AC even inside synchronized grid. Breakeven point is about 600 km for overhead lines and just 50 km for submarine cables.

And that breakeven point keeps getting shorter as the conversion equipment gets cheaper. It used to be much longer, back when the concept was new.

Solar already generates DC and inverts to the grid, so tying it directly to HVDC could probably be done by just changing the inverters, rather than adding a whole standalone conversion station. So that makes it cheaper still.

The next frontier will be when a customer asks for DC handoff.

> HVDC line is cheaper to build and has fewer losses. But you need conversion stations which are expensive and also have some loss.

I see this argument a lot lately. I fail to see why DC will be better when, at the end, you have to convert it in AC.

In a nutshell: economics. And if two grids are not in sync then you don't have many other options.

AC loses more energy via ground coupling.

Also suspect the losses for high voltage AC/DC conversion is very low. As losses tend to depend on current rather than voltage. Not exactly true because you need to connect switches in series to handle the voltage.

We've had a HVDC connection in New Zealand for a long time.

https://en.m.wikipedia.org/wiki/HVDC_Inter-Island

We’ve had it for while in the USA and Canada. Including lines built in 1932 for testing HVDC using mercury arc valves[1].

https://en.m.wikipedia.org/wiki/List_of_HVDC_projects

[1] https://en.m.wikipedia.org/wiki/Mercury-arc_valve

HVDC can be made cheaper by making the voltage higher...

As voltage goes up, the thickness of metal needed goes down. The insulator thickness goes up - that either means taller towers or thicker plastic.

Conversion from regular AC grid voltages to high voltage DC was previously very expensive, with custom made semiconductors costing millions.

With the advent of electric cars, there are now off the shelf cheap power electronics. They can be stacked, perhaps 20x in package to get to 10kV, and then 100x packages to get to 1 million volts.

That should dramatically reduce the cost of moving large amounts of power long distances.

I think that would actually be more expensive as you still have the all the hard problems. Like how do you isolate the power electronics. You can use air which is cheap but needs a lot of space. (i learned that you need 1mm per 1kV) You need a lot of space which has to be very secure. etc

Additionally you need to control 5000 converters without getting oscillations etc. (probably over glas) Also they will probably need a lot more space.

At very high voltages, I would expect the whole lot to be sealed in a polymer resin for life. 1 million volts through epoxy resin requires a spacing of just 1 inch (call it two inches for a safety margin). Then your whole converter station can be the size of a shed, and the whole project can fit within existing space on existing utility land.

How would you deal with the heat generated by the power electronics if everything is sealed in polymer resin?

Built in water cooling channels?

At these voltages, water is way too conductive.

You can use pumped oil though.

Voltage is just a potential, like gravity. Even at thousands of volts you would still need just 5 volts to operate equipment in the usual way. Communicating between large voltages can be done with fibre optic (i'm guessing) or wifi.

> As voltage goes up, the thickness of metal needed goes down. The insulator thickness goes up - that either means taller towers or thicker plastic.

Why insulate the lines at all? HVAC lines aren't insulated.

They're insulated with air... Thats why the lines have to be so high up.

As the voltage gets higher, the lines have to be higher up, and get more expensive to build. The alternative is to bury the lines, and then you use a thick layer of plastic instead. At 1 million volts, you need a layer of plastic multiple inches thick, and that starts to get heavy and expensive.

Not that simple. High voltage causes corona discharge, air is not perfectly dielectric. Reducing coronas requires spaced wire bundles for larger effective diameter. There is a sweet spot in terms of voltage above which the costs go up.

They need to be insulated against earth somehow, not necessarily by wrapping them in plastic - e.g. "taller towers".

They did it in NZ without semiconductors. In 1965 they commissioned a 600 MW link with mercury arc valves.

Ones that are close together, though.

We need about four HVDC lines through the Sierras, to connect the US wind belt (the Texas panhandle north to Canada) to the West Coast. Plus lines east to Chicago and beyond.

> Plus lines east to Chicago and beyond.

https://soogreen.com/

https://www.eenews.net/image_assets/2019/03/image_asset_5031...

Oh, good.

Now if we can only get the westbound links in place.

There's a New Mexico to Arizona link going in, which is nice for Phoenix but doesn't go far enough. The TransWest Express project (Wyoming to Las Vegas, which already has good westbound connections) has been in studies for a decade now.

I wish the Inflation Reduction Act had an eye towards funding TBM ops for these projects, more expensive in the short term but bypasses lease and right of way issues, as well as environmental studies. Climate events are going to be more extreme in the future and labor will only get more expensive due to structural demographics. “Dig Once” at this scale.

You’ve then got a large corridor for future infrastructure as well.

Honest question: I understand solar panels generate DC but I thought wind turbines generate AC. Why is a DC transmission line desirable?

It has nothing to do with generation.

HVDC is desirable because it allows non-synchronous connection of different synchronized grids.

> HVDC is desirable because it allows non-synchronous connection of different synchronized grids.

Primarily it's simply more efficient over long distances than AC.

https://en.wikipedia.org/wiki/High-voltage_direct_current#Ad...

Yeah, I forgot about the higher efficiency factor.

But not in this case - It's only a 385 mile line. If it was between two synchronous grids, it would most certainly be AC for cost/tco reasons.

Supposedly the HVDC vs HVAC break-even distance is down to something like 600-800km now[1]. Since 385 miles works out to 620km the economics are probably not all that bad. Since it is just a link between two systems, only 2 converter stations (the expensive bit for HVDC) would be required.

[1] https://www.cencepower.com/blog-posts/hvdc-transmission-syst...

Lower losses over distance, and crossing synchronization boundaries.

But this line doesn’t cross synchronization boundaries. The entire service area is synchronous.

An oversimplified explanation is that AC has less transmission losses over medium distances while high voltage DC has less transmission losses over extremely long distances. I found this stack-exchange post that goes into some of the details better[1]. It also explains some other advantages of DC.

[1] https://engineering.stackexchange.com/questions/19758/transm...

The difference is not really about distance. Your link is accurate though.

High voltage is always more efficient for power transmission over any distance than low voltage, regardless of AC vs DC, in any conductor whose resistance is not zero. This is because of I^2R losses.

For any specified value of [high] voltage and a specified wire and a specified distance, DC will be slightly more efficient than AC, for other reasons like skin effect.

But we don't want high voltage in our homes, so changing voltage down for consumption and up for transmission is essential.

Voltage changing was basically impossible with DC back when the grid was invented 120 years ago. But it was easy with AC. So our grid is AC.

Today we have power semiconductors which allow us to switch DC voltages easily. If the entire grid were being invented today -- from scratch -- it would be entirely DC.

Wikipedia tells me HVDC runs at 100s of kilovolts. What kind of semiconductor can handle that without promptly turning into plasma?

Big ones, usually made of silicon carbide.

https://en.wikipedia.org/wiki/File:Pole_2_Thyristor_Valve.jp...

My understanding is that the semi's are stacked so that any individual switch only sees a voltage it can handle, with safeguards to cut off the whole mess on failure.

I’m sure you’ve heard something like “alternating current is more efficient than direct current for transmitting power”, but this is only true in specific conditions. At very high A/C voltages, the magnetic field created by the alternating current “pushes” elections away from the core of their conductor. This has the same effect as using a smaller diameter conductor, increasing heat which increases resistance and reduces efficiency. At sufficiently high voltages and distances, DC can have less loss and also has the added bonus of not needing to consider differences in A/C frequency when transmitting power between areas that have dissimilar electrical grids.

DC is more efficient at basically any voltage. AC won because its voltage is easily stepped up and down, and high voltage is superior for long range transport vs low voltage.

But high voltage DC for huge, inter-state transport lines makes a lot of sense.

"At very high A/C voltages, the magnetic field created by the alternating current “pushes” elections away from the core of their conductor. "

Thats a frequency effect and is addressed in power-lines. The line cores are of steel cable to provide strength, but is a poor conductor. The inner core supports the Al "shell" that actually carries the current.

Or as we learned in electromag physics, why stranded wire can carry more amperage than solid (because it has more surface area).

Note: Generalizing and making a lot of assumptions. This statement is not encouragement to ignore AWG markings or limits in any way, as the effect isn't that large.

AC is never more efficient than DC. AC was easier to step up and down than DC 120 years ago. That's the only reason we use it*, despite worse efficiency.

* Motors are the other reason. Big electric motors in 1900 ran better on sinusoidal AC than DC. We have much better motor technology today, plus power semiconductors to control them. So motors are no longer a reason for the grid to be AC.

Interrupting AC is also easier than DC because DC creates sustained arcs.

AC has two problems:

1. Long lines become a "transmission line" in the sense that you get reflections. These have to Z matched and you also have to get the phase angle right.

2 Capacitance loss to ground. The lines are a massive capacitor to ground. The longer the line the more power shorts to ground

2a Because of 2. You dont use AC lines underwater. The increased dielectric means that you loose too much power from anything but the shortest runs.

You could change the frequency to a do the long runs at a lower f, and itd have its benefits, but its outweighed by the drawbacks (namely transformer saturation)

3. Skin effect becomes more of a factor. HVDC can use the whole cross section of the wire to carry power whereas AC only uses the surface. At longer distances, the material savings of thinner/lighter wires offset the higher equipmet costsat eitherend.

That's not really a problem because the current caring portion of the cable is an aluminum sheeth (good conductor, but expensive and not as strong) over a steel core (poor conductor but strong)

First some background, sorry for the long comment but I can see you need a bit of a primer before we get to your question:

Transmission lines and generators such as solar panels and wind turbines work at very different voltages. Once you need to convert the extra step to convert from AC to DC or VV isn't a really big challenge. Solar panels output anywhere from 20 to 100V, these are ganged into strings and strings are then coupled to inverters to create relatively low voltage AC, or each panel has its own inverter (not very common in solar farms). Those inverters feed into a local parallel grid which is then stepped up to join the national grid using a feed line (typically 10 to 50 KV, depending on the size of the farm and the local grid). Very large solar farms can have their own local understation where the voltage is stepped up to long haul voltage.

Sometimes there is co-generation with another source (such as solar/wind, solar/natural gas or some other combination).

https://group.vattenfall.com/press-and-media/newsroom/2019/v...

Wind Turbines usually have generators that output anywhere from 10KV to 50KV depending on the capacity and the manufacturer. This can be variable frequency current or, in a tightly grid coupled turbine it can be at the grid frequency (you can tell the difference from a distance because all of the turbines in a wind farm like that will move in lockstep with each other, this is a good indication that they are AC synchronized). At the base of every turbine you will find a an inverter and/or a step up transformer like with the solar farms. A typical turbine will do anything from 1 MW (which really is small these days, but which used to be state of the art not all that long ago) all the way up to 14 MW behemoths.

https://www.ge.com/renewableenergy/wind-energy/offshore-wind...

These are most impressive up close, to put it very mildly, think of an Eiffeltower but it rotates...

HVDC transmisison lines themselves are super high technology and you're definitely not going to find these running from every Wind Turbine to the grid, what you will most likely find is a local, intermediate AC network from a bunch of wind turbines and/or a number of solar farms to a concentration point and then a much higher voltage line from there to the national grid.

'intermediate' for shorter connections is anywhere from 10 KV to 50 KV, and for longer interconnects up to several 100 KV, all the way up to 800 KV for the longest and most power carrying lines. The engineering behind all this stuff is super impressive.

https://en.wikipedia.org/wiki/High-voltage_direct_current

Then, to answer your question:

AC suffers from something called the skin effect, it essentially means that only a small part of the cross section of a powerline carries current, effectively limiting the carrying capacity of the line to a fraction of its theoretical DC limit. So by using DC rather than AC for very long connections line losses can be minimized and much more power can be transferred through a line because those losses translate into heat generated in the line. So HVDC makes very good sense for the long haul links coupling remote areas. They might even make sense intercontinentally (though I'm a bit more skeptical about this after the pipeline attack on the NS pipelines, HVDC lines would be quite fragile and very difficult to repair after an attack).

Note that you always have these losses, but the overhead of the AC->DC->AC conversion is such that it only makes sense for longer lines or lines carrying a very large amount of power. But even the shortest AC line suffers from that skin effect.

I hope this answers your question.

As the headline says, 3 grids that are not synchronized . You cannot interconnect them using AC. It would basically be a shortcut.

Edit: You could of course keep the DC step within one site and do all lines in AC. But as the other answers say DC transmission also has benefits, so doing it inside one site you'd need all the equipment but wouldn't get all the benefits.

> As the headline says, 3 grids that are not synchronized

But why? All of the US runs 60 Hz, it should be relatively trivial to synchronize their three phases and tie the grids together. Ukraine managed to do this during an active war.

Because when they are AC-coupled (synchronized) they need to be either in balance or have energy flows no more than the interconnects can handle. Otherwise the lines will trip out in order to protect them and then both sides will be even more out of balance.

With DC you have in essence giant transistors working at each end[0] and you can control how much energy you want to flow and in which direction.

So the topology of the AC-synchronized network needs to be such that there are no bottlenecks or SPOF-ish things in the middle. With HVDC links you can build those one by one and in whichever capacity you want.

In the case of Ukraine the short answer is: there is enough transmission capacity for current operation. If there is a major event on either side and the capacity of those lines is exceed the circuit breakers will activate and separate (desynchronize) the networks. The active management of power generation capacities tries to keep the energy flows in check.

[0] by end I mean AC/DC and DC/AC conversion point, you could have more than 2 if you wanted to.

Does transatlantic HVDC make sense? Something that connects the US/Canada with mainland Europe through Greenland, Iceland, the Faroe Islands, and the UK/Norway.

Transatlantic HDVC might need actively cooled superconducting transmission lines: https://en.wikipedia.org/wiki/SuperGrid_(hydrogen)

The biggest problem with this is that the geography doesn't really make sense. A line from the east coast US to Europe goes right over Quebec which provides a large portion of the east coast's electricity through hydro already. Instead of spending billions on a trans-atlantic cable, you can just build a ton of renewables in the US and use Quebec's hydro as a giant battery.

It might be technically possible. The most ambitious project along those lines I’ve heard of is proposed from northern Australia (with lots of room for solar panels) to Singapore (which does not have a lot of room for anything). 5,000 km of HVDC undersea cable!

Not guaranteed to get off the ground, but I haven’t heard any serious reservations about the technical feasibility.

That project recently collapsed.

"Collapsed" is oversimplifying the situation.

The project was a joint venture between a multibillionaire mining magnate Andrew "Twiggy" Forrest and the Atlassian cofounder Mike Cannon-Brookes. For whatever reason Forrest decided that other renewable energy ventures were a better bet and stopped funding the project. Despite it entering administration, Cannon-Brookes gave it another $65 million to continue operations.

https://www.businessnewsaustralia.com/articles/administrator...

As such, I don't think this one is quite dead yet.

The geophysics are as sketchy as AF.

There was a certain amount of risk with the failure mode of a single deep cable etc.

That was amplified beyond acceptable by the route which crossed multiple fault lines in a heavily volcanic region with many tectonic issues.

With a tangential throw to the true father (IMHO) of evolution . . . [1]

[1] https://en.wikipedia.org/wiki/Wallace_Line

The UK is planning a HVDC connection from a wind farm near Iceland so yes, it's very likely technically possible but is it financially viable?

A similar project to this has languished for nearly a decade, if it is even still a going concern.

https://en.wikipedia.org/wiki/Tres_Amigas_SuperStation

>Skelly is behind a previous transmission development company, Clean Line Energy, which dissolved in 2017 when a multi-state project failed to secure state regulatory approval.

We desperately need the feds to take over these approvals under the interstate commerce clause. We will need a ton of transmission infrastructure over the next decade, and we can't have things being held up or denied by the states just because they happen to be in the path. See the vote in Maine over the proposed line from Canada to NY.

why should maine suffer without adequate recompense for others?

How does Maine suffer? That's not a rhetorical question; I fail to understand how a power line hurts it.

If there is such a need to put it through state land, then it can be paid to put it underground to not have to continually spray herbicides to keep the weeds down. But, that would cut into Avangrid's profits, and they have been known for substandard customer service & after storm line work. If they want to offer state level service, then the state should own the line. See https://news.ycombinator.com/item?id=29100131 108 comments & https://www.eenews.net/articles/embattled-maine-power-line-f...

I'm having trouble parsing this sentence: are you saying the federal government doesn't want to pay for the line? Or that they won't pay for some version of it?

Proposal is for the Maine line to be privately owned. Outside of the Western Area Power Administration, the generation and transmission equipment is currently owned, operated and maintained by a consortium of private industrial companies called Independent System Operators or Regional Transmission Organizations (ISOs/RTOs) administered competitively under the Federal Energy Regulatory Commission (FERC) program of the DOE. https://www.quora.com/Does-the-government-own-all-power-line...

It has nothing to do with herbicide, just some bullshit.

As with most successful rural populist movements, you follow the money. The company who owns a nuclear plant in New Hampshire funded the opposition to competition.

Long range transmission is expensive enough as it is without needing to be buried.

Forest fires are expensive too.

PG&E's problems arise from poor maintenance and mismanagement owing to their ostensibly privately run but not really ownership/governance structure.

It does forclose most other usage opportunities for the land needed.

Sure, but that doesn't equate to suffering. Presumably there's some deal that Maine would find amendable.

(Besides: Maine is has a heavily forested, virtually unpopulated interior. It's not clear to me that a HVDC line would represent a significant potential usage cost.)

The people of Maine would disagree (obviously, since they did). Forests don't exist to build over. New Englanders like their forests.

These aren’t incompatible positions: New York is both densely forested, and has an extensive interstate electricity network running through it. You wouldn’t know it unless you went looking for the lines (which I have!).

(Nothing exists “to be built over.” The only question is whether we can build a thing and whether it’s appropriate to do so.)

Have you been to the part of Maine the power line would have run through? There is nothing there. For much of the line's proposed length there weren't any small towns within dozens of miles.

...does it? The RMCC runs right through a few towns on the Quebec side. It's a power line, I don't see what the big deal is.

What about HVDC instead?

That’s what this article is talking about.

Wow I wonder if they changed the title. I swayre it said AC.

I am begging humanity to stop creating more definitions for the acronym ISO. Surely things can be described in slightly different words.

https://en.m.wikipedia.org/wiki/ISO_(disambiguation)

ISO (Independent System Operator) is a term that's been used in the energy industry for a very long time.

It's the sort of thing we should get the International Standards Organization to look into.

Funny guy.

It seems pretty clear to me that neighborhood level power grids are the way to go here. Large grids, long distance transmission and massive infrastructures are all a waste. Imagine if you could cut up a coal plant into small pieces and distribute it? Maybe a simple local neighbor grid sharing to cover a single failure.

I don't think power companies can (or want to) shift their thinking to local. Are their profits lost? Government subsidies lost?

Centralization allows massive economies of scale.

> Large grids, long distance transmission and massive infrastructures are all a waste. Imagine if you could cut up a coal plant into small pieces and distribute it?

Smaller infrastructure is actually more wasteful. Since you suggested a coal plant, consider the sheer number of extra jobs it would take to hire people to drive the coal truck to the neighborhood plant and handle the offloading of coal. At the scale of per-neighborhood, you might even need someone to literally shovel coal into a furnace, I don't know if the coal conveyor lines make sense at that tiny a scale. Double that because you need to transport the resulting fly ash as well. Then there's the extra instrumentation you need, the duplicate emissions mitigation points. And this is before considering the literal better efficiency of larger boilers/blowers/etc. themselves.

In a concrete example I have familiarity with: I worked for a large water treatment plant that supplied water for about a million people. It required 2 engineers being present 24/7/365 to keep it running. Replace it with a water treatment plant that is 1/100th the size, and you still need 2 engineers present 24/7/365. Go much smaller than that, and the inability to economically provide on-site permanent engineer oversight leads to the idea of controlling the water treatment plant remotely via the public internet, with all of the increased risks that entails.

The other half of the micro hyper-local grid is to get out of the business of needing very large plants like coal plants and moving to things like distributed wind and solar (perhaps as local as being on every rooftop in town) with battery storage

That's of course easier said than done, but with things like wind/solar/battery, and small failsafe nuclear you can get to smaller grids that don't have to be so absolutely interconnected and trying to ship power from Nebraska to New York City

The problem with distributed wind and solar is that some places are better than others. Solar can go on roofs everywhere but it works better in sunny places. Utility solar is cheaper than roofs. Wind has places where it is stronger like offshore and the plains. Also, windmills are more efficient when bigger.

The result is that it is more efficient to put wind and solar in best places and ship the power to where needed. The independent survival area probably depends on where end up putting batteries. I suspect that substations and neighborhoods make most sense but houses might also be best.

Both your examples are reasons for the consolidation of existing infrastructure. My point was that you couldn't break up a coal plant, but with renewables they are already broken up into small pieces. Putting the pieces together at one location instead of distributing them as they exist naturally seems like a poor design choice to me. And I wonder if it is motivated by the fear of loss of control and money more than just a poor decision?

Seems like hvdc would be great across hemispheres so summer solar can go to winter areas.

It is a lot cheaper to build transmission lines compared to providing storage.

In the future the vast majority of electricity will come from wind and solar. If production does not match consumption then it is almost always the best option to move the electricity somewhere else.

(That said: Of course storage will be needed, but only if there are no better options to balance the grid otherwise.)

Reply on news.ycombinator.com