The Baltic states are reconfiguring their electric grids | Hacker News Reader

The Baltic states are reconfiguring their electric grids

economist.com

135 pointssohkamyung77 comments

This article is completely off point. The three Baltic states have inherited Soviet power system and back then they produced more power than they needed and they were net power exporters. However, in order to join the EU, they had to comply with very strict environmental regulations. For instance, Lithuania had to shut down Ignalina nuclear power plant, this was a required condition from the EU or Lithuania wouldn't be able to join the Union. Last year Estonia had to decommission their shale plants (yes, they used to burn shale) due to CO2. In order to get power from somewhere, a series of interconnectors were built, with Sweden and Finland (which is funny, 40% of Sweden's power comes from nuclear) and those interconnectors had 16 outages within a very short period of time, which happens when you build new grids, you need to work all kinks out. As the result of this, the price of power in the Baltic states increases by about 30% each year. So, this is the background story that you need to understand before reading the article. If you first destroy your power generation capacity and become depended on imports - yeah, you become vulnerable. Excess capacity is unprofitable BUT it's much safer to have it. Right now excess capacity for the Balitc states is Russia and they want to change it to the EU. But physics doesn't care about your political views. It's always smart to have internal excess capacity rather than depend on outside parties.

Your post is mostly right. I haven't read the article itself (paywall) but I already knew about this particular issue through work.

You've touched on a lot of the "pull" factors that cause the Baltic states to connect to the EU. There are however a few geopolitical "push" factors that cause Baltic states (particularly Lithuania) from staying dependent upon connections to Russia.

The Lithuanian grid has been able to rely on energy from Russia due to the Kaliningrad enclave being dependent upon the Lithuanian grid. Russia has recently been making investments into making Kaliningrad energy independent. If this is achieved, Russia no longer has a vested interest in the security of the Lithuanian grid.

Given the history of alleged Russian sabotage of Lithuanian infrastructure (other than electrical), there is considerable fear that once Kaliningrad is energy independent that Russia may sabotage the Lithuanian grid. Therefore Lithuania is rushing to connect to the EU grid before Kaliningrad achieves self-sufficiency.

> Lithuania had to shut down Ignalina nuclear power plant, this was a required condition from the EU or Lithuania wouldn't be able to join the Union

> which is funny, 40% of Sweden's power comes from nuclear

I think the issue there was that this was RBMK-type reactors, which is the same type as in Chernobyl, and it is generally considered unsafe, even with the safety upgrades that was done to it.

I believe HBO Chernobyl used Ignalina as a filming location.

Operating a reactor designed for weapon grade fissile material production (which RBMK is) does not look too good, either.

“Megatons to megawatts“ was a good slogan though. Agreed it’s not the best approach.

You are correct. And there we talks to build a new nuclear plan for all three Baltic states but they don't seem to have any progress. Moreover, Belarus has built a new nuclear plant 30 miles away from the Lithuanian border (the first reactor is being loaded as we speak) and Lithuania has actually banned any purchases from that plant, while Latvia and Estonia have not. So, there's a lot of politics around nuclear power there as well.

The problem with Ignalina was not that it was nuclear, it was that it used a horrible old design that both can and have blown up in catastrophic ways. Note that Sweden is also slowly decommissioning our old bad plants.

Another interesting recent reminder of the interconnectedness of politics and electricity grids happened a couple of years ago, when all of Europe's frequency (and lots of grid-synced clocks) went out of spec because of a Kosovo/Serbian spat: https://www.theverge.com/2018/3/8/17095440/europe-clocks-run...

This is so cool! Incoming rabbit hole of electric grid research...

Interesting. I don't understand everything so I have a question to the experts here: It seems that a small part of Kosovo stopped paying for electricity, so the electricity producers stopped producing for the parts. However this didn't work out, electricity still flowed to the nonpayants, because it seems that nobody cut the wires. This created an inbalance distorting the grid frequency, such that clocks in Europe ran a bit late (6 minutes). Sort of an Europe-wide soft brown-out. It's not told in the article, but I think the voltage was also a bit lower than it should be.

Did I understand this correctly?

It is pretty much right.

Neither country took responsibility for the demand (X), and so the market wasn't aware of the demand, and therefore the generation provided by the market was lacking X. When there is more energy being taken out of the grid than being provided, generators start to slow down. That is, the power is extracted from their rotational energy (inertia), meaning their rotational speed slows down (frequency decreases).

There are mechanisms in place that fix any frequency deviations, but the assumption is that deviations in frequency are symmetrically distributed around 50 Hz (or 60 Hz for you americans). When you have consistently skewed errors (i.e. frequency always being pulled down by Kosovo), you end up with a 'time error' [0].

The time error was noticed fairly quickly (frequency and analogs to 'time error' are monitored in control rooms), but the issue was that no-one took responsibility for it and because it became a political issue it was difficult for other grid operators to step in and fix it.

I'm not sure about voltage being lower in this case. Voltage is a very local phenomena, and it would be very unusual for control rooms to allow voltages to be 'lower than they should be', especially over a considerable time period. To do so they would be in violation of their grid codes.

[0] https://www.nerc.com/docs/oc/rs/NERC%20Balancing%20and%20Fre...

> A giant outage in Latvia on June ninth concentrated minds

"Giant" per what definition? It happened because of damaged isolation for 330kV line and automatics just shut it off. It also brought overflow within 110kV line affecting 160k people in capital and around it.

Communications was affected (apparantly they didnt test their energy backup systems), traffic, homes. Diesel generators were put to work where needed. Our datacenter was up and running, being backed up by other electricity lines and diesel.

Yeah, a MAJOR outage, but in LESS than 2 hours everything was restored.

Nothing to do with frequency. Why even mention it and try to associate with Russia?

Russia is an enormous threat to Western civilization, haven't you heard? /s

>> Nothing to do with frequency. Why even mention it and try to associate with Russia?

> Russia is an enormous threat to Western civilization, haven't you heard? /s

Merely yet more anti-russian neocon propaganda, who are nostalgic for the cold war. It's the $insert_one_of: Chinese, Russian, North Korean, Iranian hackers who took down the electric grid, hacked hospitals etc.

It's not like Russian agents go around attacking The West, doing things like interfering in elections or launching chemical warfare in Western nations is it?

Oh, wait, yes - that's exactly what they've done.

Chemical warfare?

Meaning assassination with polonium, I guess. I dunno if it's chemical warfare but not something I expect from a reasonable government.

That too, but I was referring to the Novichok poisonings in 2018: https://en.wikipedia.org/wiki/Poisoning_of_Sergei_and_Yulia_...

I think this is referring to the poisonings in Salisbury in England in 2018.

Ah, Cold War... Probably, these two Russian agents who committed an assasination attempt (led to innocent human casualties), using chemical weapon on the soil of a NATO country, were particularly nostalgic.

Sharing a land border with Russia also tends to concentrate minds

Grid operators have many ways to measure the severity of an outage. In Europe, this is commonly done using the Incident Classification Scale [0].

The term "giant" is not any official terminology, but that particular outage would be considered an "Extensive Incident" (red) due to it being in the T2 category, ON2 category, and possibly in the L2 category (maybe L1, I don't know the Latvian grid well enough).

A lot of people in the industry would say that a 'red' incident is 'giant', as they tend to have high standards for reliability and security.

[0] - https://eepublicdownloads.blob.core.windows.net/public-cdn-c...

Just a question:

Why does demand exceeding supply cause the frequency to drop?

In my mind it should cause the voltage to drop?

If this question is too long to answer on HN, are there any good resources that explain this?

Well, if anyone sees my comment, I found some (I think) great answers on Reddit. So if anyone else has the same question here is (as far as my non-electrical engineer mind can tell) some great answers.

https://old.reddit.com/r/AskEngineers/comments/1tnbob/why_do...

This is a gross simplification but hopefully makes sense:

Grids using multiple generators use "frequency droop," deliberately governing the generators so that as power goes up, frequency goes down. This makes it easier to load-balance between generators connected in parallel.

If one generator were to be "overloaded," its frequency would go down, increasing power output from other generators.

When connecting a new machine, it's frequency is typically slightly higher than the grid; after it is connected it "assumes" some load and its frequency drops to match the grid's frequency.

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

And to simplify even more:

Generators are just spinning magnets. When load goes up, magnetic resistance goes up, so given the same input mechanical power, the spinning will slow down.

A coal power plant in Thunder Bay, Ontario used to drive an otherwise decommissioned generator for the sake of having inertial mass to stabilize the frequency of the local supply.

I would imagine this is common practice in power generation in general.

Yes, it is common practice:

https://business.directenergy.com/blog/2016/may/reactive-pow... https://openei.org/wiki/Definition:Reactive_Power

Yes, it's called a "Synchronous Condenser"

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

In an AC grid, increased or decreased load affects resistance which impacts the rate at which the mechanical source of power generation can rotate and therefore the frequency of the current:

https://physics.stackexchange.com/a/2661

If you add load to a generator, it increases the amount of force needed to turn it. (There's a nice experiment with a hand crank generator that has a switched load - it's very noticeable, and logical - the electrical power you use needs to come from somewhere).

If the amount of force that's available doesn't increase, the generator slows down. The generators are synchronized to the grid. Of course, as the frequency drops, power plants regulate by applying more force to the generators (e.g. increasing the amount of gas or water fed to a turbine).

If you've a small load dynamo spins faster as there is much less force required to make it spin.

So frequency should go up, not down.

As the spin is directly proportional to the frequency

GP comment: Why does increasing load make the frequency go down?

Parent comment: Increasing load makes the frequency go down because ...

Your comment: Decreasing load makes the frequency go up

Your comment is in agreement with everything before, but you said it in a tone of disagreement. Maybe you misread the parent comment?

ah, yes i did. thanks :)

I first came across this phenomenon with the Kosovo issue mentioned elsewhere. But it immediately made sense to me from driving a gas car (you don’t notice it in an electric): hold your foot on the gas pedal without moving it and go up a hill. What happens? The engine RPM go down and also the car slows down (because it depends on RPM to the wheels). As the load increases, without more power (fuel flow) the engine slows down.

Same with an electric generator/motor: if you pull power off more than it’s being driven, each push from the driving field is less than enough for the rotational rate, so the rotation is slower, and thus the frequency too.

Interesting that the US grid is not national, according to https://en.wikipedia.org/wiki/File:Wide_area_synchronous_gri...

Am I misunderstanding "synchronous"? Would it really be possible to mis-frequency only part of the IPS/UPS grid without bricking the rest?

Combined, the Baltic States are about 67k square miles. This is smaller than Missouri. Smaller than nearly half of the US states. Alaska by itself is ten times their area. And though you might say, "but that's Alaska" that's exactly why comparing the US to European nations at the national level is usually a category error. Individually, the Baltics would rank in the smallest ten if they were US states.

Or to put it another way, for things like utility grids and high speed rail, a national level system in the US is comparable to a pan-european system that includes everything from Lisbon, to Oslo, to Bucharest and Belgrade and Istambul.

> Or to put it another way, for things like utility grids and high speed rail, a national level system in the US is comparable to a pan-european system that includes everything from Lisbon, to Oslo, to Bucharest and Belgrade and Istambul.

Which is exactly what exists, and what the Baltic states are joining: https://en.m.wikipedia.org/wiki/Synchronous_grid_of_Continen...

It doesn't go as far north as Oslo, but does include Morocco, Algeria and Tunisia.

>> Smaller than nearly half of the US states. Alaska by itself is ten times their area.

I wonder if there is something between those states, maybe some sort of land bridge linking Alaska to the more southern states? The north american grid is very multinational, with the US generally a net importer of Canadian power, especially in the east. Except texas. They are weird.

https://en.wikipedia.org/wiki/File:NERC-map-en.svg

(And be careful when talking about the size of US states. Americans don't like being reminded how small they are.)

Functionally speaking, the infrastructure of Canada doesn't really expand into the northern territories. The Alaska Highway was only built as an emergency measure during WW2; without it, there'd be no infrastructure connection between (non-panhandle) Alaska and British Columbia, or even much of the Yukon and British Columbia.

> (And be careful when talking about the size of US states. Americans don't like being reminded how small they are.)

The US actually has slightly more land area than Canada, as a larger fraction of Canada's area is large lakes such as Lake Winnipeg, Lake Manitoba, Great Slave Lake, etc.

Canada has built many highways since WWII, including a second, straighter, highway through the yukon to alaska.

https://upload.wikimedia.org/wikipedia/commons/9/9b/Roads-Ca...

The alaska panhandle and BC coast are a series of islands not suitable for highways, or power lines, or much else. As a practical matter, it isn't really connected to canada. Much of it is effectively more remote than canada's northern territories.

This is considered a feature, because having multiple independent systems reduces the probability of a complete failure. IIRC, some national security-sensitive facilities are duplicated on two grids.

The really weird thing is that Texas has its own grid. Especially since electricity is significantly cheaper on the Texas grid than on the other two.

Isn't it that chokepoints are what causes failures in the electric grid, not general connectedness? If you are looking at security holes in control equipment it doesn't matter of you are connected to a grid, when you've got the NSA/KGB/Chinese Ministry of State Security inside your generating station you've lost anyway.

Failures can cascade across the grid. A software bug in an Ohio utility's computers in 2003 ultimately took out power for 55 million people in several US states and the province of Ontario. https://en.wikipedia.org/wiki/Northeast_blackout_of_2003

If you have sites in multiple grids you're guaranteed that one failure doesn't take down every site.

Never heard it was a software bug? Lack of maintenance allowed tree branches to grow under transmission lines which sag the hotter they get and so on a hot summer day with record system loads one transmission line tripping caused a cascading outage.

Read the Wikipedia page, sure the system operators in MISO didn’t get some alarms but blaming the whole thing on that is pretty lame. There were multiple causes and not Getting alarms was just one of 10 things that went wrong.

we do what we want! which is wind power somehow

It's not weird at all. Texas disconnected from the East and the West via AC ties (still has DC ties that connect to the Eastern and Western Interconnects) to avoid FERC's federal oversight, so they only answer to the Texas PUC.

Electricity is cheaper there due to the prevalence of wind in comparison to much of the country, but there are also really low prices in the SPP region (includes part of Texas, Oklahoma, Kansas, Nebraska, and parts of several other states) due to wind.

It is also important to remember that ERCOT does not have a capacity market like every other FERC regulated ISO. The added cost in places like PJM and CAISO is noticeable. Though having capacity markets does not seem to prevent blackouts.

In practice, capacity markets seem to be mostly handouts to incumbents with a lot of fossil capacity.

ERCOT has an ORDC, further pushing up prices during a crunch. OTOH the rest of the time consumers enjoy quite low prices.

Yep $9000 prices upon occasion.

That is not entirely true. SPP does not have a capacity market as capacity procurement (outside of resource adequacy) seems to be on the States.

Yes, North America is not a single synchronous grid. All of it operates at 60Hz, but interconnections between the regions are done using either HVDC or VFTs because electrical signals from both regions would not be in-phase (i.e. have synchronized peaks)

https://en.wikipedia.org/wiki/Western_Interconnection has details on how the Western Interconnection works.

I wonder what the actual advantage of a synchronous grid is.

You can transfer power from one side to another without the extra hardware expense and losses that a HVDC connection imposes.

Non-synchronous interconnects are more expensive, inefficient, and complex than synchronous ones.

Historically cost. Building a non-synchronous grid is much harder because you need equipment that can correct for phase differences in your grid. This equipment used to be very large and very expensive, Mercury-arc valves are some of equipment that used to be used (they provide AC to DC conversion) [1].

But modern power electronics are so cheap and efficient that High Voltage DC transmission lines are becoming increasingly common, driven by the fact the High Voltage DC is more energy efficient that High Voltage AC (you can push more power down the same cable, with less energy lost to resistive loads because DC doesn't suffer from the skin effect [2]).

With more High Voltage DC transmissions lines it becomes easier to run non-synchronous grids, because the DC transmissions acts as natural breaks between synchronous grids. Running a non-synchronous grid tends to be a little easier because you no longer need to worry about strange phase variances caused by multiple transmission lines of different lengths injecting phase delays into your network.

[1]: https://en.wikipedia.org/wiki/Mercury-arc_valve [2]: https://en.wikipedia.org/wiki/Skin_effect

That's sort of what I was thinking. Seems easier to coordinate a HVDC interties than a synchronous grid. I also keep coming back to the thought that long east west and north south interties would be beneficial when you have large percentages of solar and wind.

You can provide frequency control in one part to keep another part stable. For example if you have a part of your grid with a high level of renewables (for example, South Australia), system stability can be provided by large intertia generators in other parts of the grid (in our example, coal power stations in Victoria).

Take away the interconnect, and the islanded part of the grid is no longer stable, requiring the renewables to be switched off and replaced with gas and diesel generators. This happened after a storm earlier this year:

https://www.afr.com/companies/energy/sa-to-remain-cut-off-fr...

If two interface points take different-length paths from the origin, they might be out of phase at the interfaces. If I'm doing the math right, a path 250 miles longer would be 90 degrees out of phase.

There are many factors that affect the phase angle. Path length is one, so is the impedance of the lines (a line with double the X will get the same phase offset in half the length). Other equipment will have phase offsets such as transformers where the phase offset is a function of the winding configuration. There are event specialized transformers just for this purpose https://en.wikipedia.org/wiki/Quadrature_booster

The (very real [1]) risk is that the Russians could disconnect the Baltic states from IPS/UPS, forcing them to operate autonomously, at a time of high demand and low local power generation. This could rapidly (within seconds) cause a grid-wise blackout which could take days to recover from - you have to do what's called a "black start" [2] to restore service - which is technically quite challenging.

[1] https://core.ac.uk/download/pdf/38632048.pdf [2] https://en.m.wikipedia.org/wiki/Black_start

Importantly, from your first citation:

> The present report gives an insight in the ability of the Baltic States to operate their electricity systems independently of the neighbouring countries. The results show that by 2020 and 2030 the dependency of the Baltic States on the outside generation resources is fairly low, following current electricity system development trends.

Consequently, if one were to desire to disempower Russia and Belarus, you’d increase funding to the Baltic states to allow them a more rapid transition to robust electrical grids (through transmission, generation, and interconnector upgrades). The Baltics could always get more power from Finland and Sweden and disconnect entirely from Russia and Belarus if necessary.

> The Baltics could always get more power from Finland and Sweden and disconnect entirely from Russia and Belarus if necessary.

Estonia has been a net exporter of electricity since 2017. [1] Amusingly one of the export targets is Russia.

Looking at the current live data [2] it looks like Latvia is in a similar situation, although I haven't searched for yearly statistics.

Lithuania is a net importer though, but that's largely due to major restructuring of its electricity system. Specifically because a condition of Lithuania joining the EU was that it has to shut down its nuclear power plants. If those plants were still running then Lithuania would be a big exporter, because the plants were built during Soviet times with the idea of powering a bunch of the neighboring coutries as well. [3][4]

[1] https://majandus24.postimees.ee/6738108/energiasoltumatus-on...

[2] https://www.electricitymap.org/zone/LV

[3] https://en.wikipedia.org/wiki/Nuclear_energy_in_Lithuania

[4] https://en.wikipedia.org/wiki/Ignalina_Nuclear_Power_Plant

The EU required Lithuania to shut the nuclear reactor in question (Ignalina) down because it had an inherently flawed design making it unsafe to continue to operate. There is still an opportunity to use storage and renewables to be a power exporter. The renewables will need to be built regardless.

I highly doubt that the Baltic countries will exist in 2030, rather they'll be already annexed by Russia(this will happen around the middle of this decade).

I don't have high hopes in NATO, but I really doubt that would happen

I'm afraid its going this way. Belarus will join Russia this year. Trump will get reelected and throw the Baltics to Russia in exchange of the new nuclear detente.

NATO and the EU are useless - NATO depends on the US, EU is toothless and their biggest military power - France is Russian slut.

The only solution would be to show strength NOW - deploy nuclear weapons to the Baltics and deal with every single fifth columnist there in a brutal, public fashion. Of course this will never ever happen. Annexation is matter of time.

I'm moving away from Tallinn next year, so hopefully this year nothing will happen...just in case I have all my money moved out of the country.

Russia doesn't exist for centuries. Great Russian Empire doesn't exist for century. Russian Federation may not exist in 2030.

This is the most entertaining (and uninformed) statement I've read today. I'd be curious to know what makes you think this is a plausible scenario?

Looking at the map, reading the news, reading history books. Basically summing up 2 and 2 together.

Another thing that's not mentioned in the article is the Astravets nuclear power plant in Belarus, which is largely subsidized by Russian funds and is seen as a political maneuver by Russia to threaten Lithuanian sovereignty. Lithuanian energy situation has been an issue ever since the closing Ignalina nuclear power plant as a condition of joining the European union. There were plans to build a new modern nuclear plant in its place, but it has been canceled, largely as a consequence of the 2008 financial crisis. The Astravets power plant is seen as a way to force Lithuania to buy cheap electricity from Belarus and keep close ties to the Russian sphere of influence. It's choice of location is largely questionable, being built just 50km from Vilnius, the capital of Lithuania. The power plant is set to start generating electricity sometime this year and Lithuania is looking to reconfigure its power grid in time to boycott it.

This reminded me of the very interesting (to me) report and discussion on the failure of the UK national grid last August: https://news.ycombinator.com/item?id=20973513

http://archive.is/B1Qx8

I couldn't read most of the original article because paywall. This https://www.onlinefinancenews.com/why-the-baltic-states-are-... is not behind a paywall. It also seems to be (machine?) translated.

The Economist paywall can be bypassed by disabling Javascript on the browser. They used to have a limited number of articles you could read per week for free, but it looks like that's not the case anymore. It's really a shame because their journalism is really high quality, but I guess they do have to pay for it somehow.

Reply on news.ycombinator.com