Hacker News Reader

> "solar generates DC but wind generates AC."

Yes, but a wind turbine is allowed to spin at variable speeds - its rotation is not synchronized to the grid frequency in the same way that hydroelectric and thermal turbines are.

In order to get a wind turbine's power output to match the grid frequency, it goes through an AC -> DC -> AC conversion in a component known as a double-fed induction generator (DFIG).

Wind turbines, the big ones, regulate themselves to turn in sync with the grid. The adjust thier blades as needed to keep pace/power. That's why huge fields of them all turn in perfect lockstep.

From some random reading on the net, it is the smaller ones that can be directly connected. Above 2MW things are not as easy and it is better and more efficient to convert to DC.

Wind turbines have frequency converters. They are semiconductor devices. In one 3 MW turbine I visited, it was at the base of the tower.

The blade angle is adjusted to get optimum power extraction. Rotor RPM is completely independent of produced power frequency.

References: https://www.vestas.com/en/products/offshore/V236-15MW and https://library.e.abb.com/public/bf09cdf11d234241845c79ac343...

You're talking past each other.

The electronics perform two key functions:

* Choosing pitch angles for efficiency and turbine safety. You can, for smaller turbines, just synchronize the turbine to the grid, but this is becoming uncommon practice.

* Converting the produced AC power to DC, and then choosing the proper frequency output and voltage to feed power to the grid, and inverting the DC to make this power. This should usually be trying to "speed up" the frequency of the grid a little if it's not already way too fast and regulate the voltage appropriately.

The second link you have, on page 3, shows (active) rectification (d) of the wind turbine AC power to make positive and negative DC buses, and then inversion of that DC (f) to make 3 phase AC output power.

Yeah, I don't know if for example in big wind offshore wind parks, since it's connected with a DC cable to mainland anyway, you would convert all the turbine generator outputs just to DC.

Then at the ground station where it's connected to the grid, convert to AC, whatever kilovolts are needed.

> you would convert all the turbine generator outputs just to DC.

You could. But you probably still need a DC/DC conversion step or boosting in order to let the power flow from each of the turbines fairly. (The synchronous conversion to DC from AC provides opportunity to slightly change the voltages you get out, but not terrifically so).

Blade pitching first use is to reduce the aerodynamics power above the rater power point of the turbine and avoid overload.

Yup, that's why the word "safety" is in there. I probably could have put that better. (It also includes e.g. feathering rather than just trying to stay close to rated power, speed, and torque limits)

They are still not syncing directly to the AC grid. They are not going 50 rounds per second

They wouldn't necessarily need to do 50 rounds per second to generate 50 Hz. There could be any number of poles on the generator. At least in theory... Not saying you're wrong about not being synced but just that they wouldn't need to do 50 rps...

Gearboxes. Windmills of the same design will turn at the same rate in lockstep with each other. A gearbox converts that rate into what is necessary for the generator. The generator turns in sync with the grid.

> "That's why huge fields of them all turn in perfect lockstep."

Not true for any modern wind farm. While they might appear to be turning together, that's just because the wind conditions are relatively consistent across the whole area. Not because they are actually synchronised! Most modern turbines will target a certain rated/optimal/maximum rpm once a certain wind speed has been reached, but are free to rotate more slowly (while still generating power) in light wind conditions.

Likewise for solar -- the DC voltage and current a panel puts out is dependent on both insolation and load -- only in the most simple applications can you just hook in directly to a panel, without involving a DC-DC converter. Even grid-tie installations often use DC-DC converters, called "power optimizers" [1], to optimally tie multiple panels to a single bus.

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

I have personally worked on a tens-of-kW-scale wind turbine which was directly connected to the grid and was fixed speed and blade angle. It's quite common at lower power ratings because the cost/complexity of pitch control isn't justifiable.

It wasn't much more complicated than a three phase electric motor bolted to high-angle-of-attack blades, connected to the grid via contactors controlled by a microcontroller and a grid-intertie monitor.

The author seems very confused about whether they're talking about the grid or devices.

Which is surprising, given his background, a degree in electrical engineering and jobs with power companies.[1]

This may be an argument for using more DC-DC converters and fewer transformers. The classic problem with shipping DC around is that voltage conversion is expensive. DC-DC converters have improved a lot. This article may be a dumbed-down version of that argument.

A nice thing about large transformers is that those big hunks of copper and iron have a lifespan of 30 to 75 years. Replace those with a DC-DC converter, and it will probably have semiconductor lifespan problems. Plus someone will add on a data connection, firmware updates, a web server, and an antivirus program.

[1] https://grid.pitt.edu/people/gregory-f-reed

> "DC power is significantly more energy efficient than AC power." -> the examples go on to specify end points for electrical energy but we already use DC there, AC is mainly used in transmission, so the claimed advantages of DC are irrelevant.

DC is more efficient for transmission too.

The reason the grid uses AC is that high voltage is always[1] more efficient for transmission than low voltage--regardless of whether you're talking about AC or DC--and in 1900 the only way to step voltages up or down was to use transformers, and transformers require AC. That's unfortunate because for a given voltage, it's more efficient to transmit it as DC than AC.

But today we have power electronics which can step DC voltages up or down, which means we are free to convert the grid to DC.

[1] The single exception is superconducting cables which can transmit low voltage power just as efficiently as high voltage power. But they are not yet cheap enough to be practical except over short distances.

I fully agree with what you said about AC/DC, just want to point out that DC benefits superconductors as well since superconductors have critical current densities where they lose their super conductivity.

In terms of motors, yes -- virtually all motors need to provide AC to the coils to run, so DC motors need to use an inverter.

But increasingly these days, even AC motors are being run from variable-frequency drives, in order to squeeze out a bit more efficiency, because the savings from better matching the load more than makes up for the losses in the drive. Many jurisdictions are starting to incentivize or require VFDs for HVAC applications. And typically the first thing the VFD does is rectify the AC input to DC.

In marketing speak, a variable frequency drive is often referred to as "inverter technology" in appliances.

Indeed. There seems to be a lot of confusion about the types of AC. "AC" can mean "50 or 60 Hz sinusoid synchronized to the grid" or it can mean "a waveform that is not a constant voltage." Most motors require some form of the latter. Older motors required the former, but there are very few reasons to build motors that way today. The ones that remain in service can easily be driven by electronic inverters from a DC supply.

DC is generically more efficient for transmission than AC. That is why it is used for very long lines. It used to be that you needed to stay AC to use transformers to step up the voltage but those days are long past.

>Brushless DC motors actually require a separate circuit to turn DC into something resembling a sinusoidal current (i.e. AC).

Typical split phase AC induction motors used in residential applications are not very efficient and have various other deficiencies. There is a tendency to do a AC>DC>AC thing to a 3 phase these days for smaller electric motors and get variable speed as a bonus.

>...wind generates AC.

But not at any particular frequency. So typical wind turbines have a AC>DC>AC converter to allow them to sync up with the grid.

If you wanted to efficiently make an efficient deeply variable speed modern motor run on AC, you might very well turn the AC into DC then back into AC in a brushless motor controller.

Indeed, changing the frequency of an AC wave programmatically is incredibly difficult (but possible through a CVT, I guess), you're better off turning it into DC then back into AC through some some form of function generator.

So indeed, the widespread use of BLDC motors is a point in favour of DC electric circuits in the home.

Same goes for variable velocity generators, you will generally have an AC-DC-AC conversion in a variable speed generator. Either that, or a gearbox, those are your two options.

Youden, your comment is a great example of why I love HN so much. Thanks for sharing this, and clearing a few important things about the article.

The case for DC is named microgrids: keeping the frequency on a microgrid is hard, inverters react too slowly for most loads (that does not ramp up/does down power slowly) while DC devices can be far quicker. The case for AC is large grid and safety.

AFAIK wind turbines, as opposed to conventional turbines in power plants, don't directly connect to the 60Hz Grid but go through DC and an inverter. This is done so they can efficiently work at different speeds at not just a few mechanically selectable ones

They have to, because the variable wind means that they need to be spinnable at a continuous range of speeds.

Components connected to the AC grid need to synchronize with the grid's frequency. Since we can't force the wind to blow at a particular rate, we'd either need a lot of fancy mechanics on the turbines themselves to drop their speed (which would waste energy) or we decouple their spin rate from the grid frequency with the AC -> DC -> AC converter (which also wastes energy, but probably less and with much less cost than complicated spin-rate-stabilizing machinery).

You could also have a continuously variable gearbox, which is still hugely problematic, but would waste less energy.