Rimac unveils SineStack, battery energy storage system | Hacker News Reader

Rimac unveils SineStack, battery energy storage system

energy-storage.news

116 pointstaubek74 comments

In the world of EVs, top auto makers are forced to become experts in batteries and battery management systems.

The record-breaking Rimac Nevera can pull up to 1.4 megawatt from its battery (https://www.rimac-newsroom.com/press-releases/rimac-automobi...)

Even my lowly leaf quickcharges at 50KW. The traction is 80KW(plus losses). That is already a lot of power.

Compared to a house, that's a lot. Not even a modern house on a 200A breaker supplying 220V can go that high - could get around 44KW but not continuously. Many homes are still on 100A service.

1.4MW on a car is mind-blowing and in the realm of sci-fi not too long ago.

This tech will eventually become mundane and available everywhere (and hopefully some of that will leak into other areas). But for now it's still spectacular.

> This tech will eventually become mundane and available everywhere

It probably won't, due to sheer physics, and the cost of dealing with them. However, what it will do is drop the price floor out of lower power applications - no longer will that level of tech be competing for the high end. It's exciting!

I've been pretty impressed with the Leaf - I don't have one, but some friends do. Pretty hackable and serviceable on your own. The acceleration and torque on that thing is insane. It's pretty easy to swap out the batteries of an early generation with a newer generation, and even possible to build your own out of a stack of 18650s. If only the integrated frame was as easily serviceable :)

If I had a use for such a car, I would get one - (perhaps) unfortunately, my driving needs are limited to going larger distances, often with off-road stretches.

1.4 MW is subway train territory.

I wonder how long that lasts.

Not sure if enough people have Rimac on the radar. They are pushing hard. For example the new Bugatti cars will be made by a joint venture of Rimac and Bugatti (=Volkswagen), and they will be hybrid or all electric. They have their own hyper car (the Nevera which accelerates from 0 to 60mph in 1.74 seconds), do development work for other car companies (like Porsche) and now present this BESS. This is exciting stuff coming from Croatia.

I'm not sure framing it as a joint venture is the right way to frame it, though they use the term to describe it that way. VAG put Mate in charge of a new division - Bugatti Rimac Automobiles. They are the parent company of Bugatti and Rimac. In it, Rimac holds 55%, VAG owns 45%. However, VAG also owns 22% of Rimac below, so VAG has technical controlling interest here. By any regards though, they basically did an Apple/NEXT type of takeover, where VAG bought much of Rimac but Mate took over Bugatti. Obviously, I think this all makes your point stronger.

They didn't buy Rimac, they bought a stake. It used to be 24%, not sure if it's bigger now https://www.caranddriver.com/news/a36920070/bugatti-rimac-po...

I updated the comment, It's really weird. They bought a part but also effectively gave Rimac the Bugatti company. One can argue that they basically leveraged Bugatti to acquire some control of Rimac and board seats without spending actual money.

They have a glorious flowchart on their site. https://web-cdn.rimac-automobili.com/wp-content/uploads/2021...

That IS a glorious chart, thank you; am I the only one who would REALLY want some arrows there? :)

It would be hilarious if it also had the fact that Porsche is owned by Volkswagen, and Volkswagen in turn... owns Porsche. https://www.bosshunting.com.au/motors/cars/volkswagen-group-...

That chart reminds me of Korean chaebols ownership structure right before the big crash of 1997

Asianometry How the Rich Ate South Korea https://youtu.be/hWCcvOE84Ao?t=489

Porsche has a 45% stake in Bugatti Rimac and a 22% stake in Rimac Group, which has a 55% stake in Bugatti Rimac. So it's complicated.

I like my conglomerates like my datastructures, directed and acyclic :p

Not that complicated. Porsche was once described as "the hedge fund that happens to also build cars".

Porsche did this stake here and there thing precisely to gain > 50% of Rimac voting rights.

It's simple really: Porsche controls Rimac.

Then there's the whole "VW took control of Porsche after the short squeeze made by Porsche to try to acquire VW failed" (it failed due to the sudden crash during the 2008 banking crisis IIRC). And now Porsche is apparently spinning out of VW and shall be independent again.

They only own 22% in Rimac, that's not a controlling stake. Therefore they don't have > 50% of Rimac (Rimac Bugatti I assume) voting rights.

I missed where VAG bought Rimac. Are you sure this is true? Here is an org chart for Rimac Bugatti, and AFAIK the ownership structure has not changed since:

https://www.rimac-automobili.com/media/press-releases/rimac-...

It's not Volkswagen AG who took control of Bugatti and Rimac, it's Porsche.

And Porsche is spinning out of VW and becoming independent again:

https://capital.com/porsche-volkswagen-vw-demerger-spin-off

https://www.autoweek.com/news/green-cars/a36933399/did-porsc...

(answer is "yes" btw)

Huh, why would VAG just give up on Porsche like that? Especially that Porsche holdings owns a big part of VAG itself.

Because the people controlling VW are the same that arw behind Porsche, namely the families Porsche and Piech.

Ah but Porsche is not Porsche but instead Porsche AG and Porsche SE.

It is pure insanity in need of many flow charts.

Independent, except VW AG still holds a majority of the ordinary voting shares and the rest of the Porsche AG voting shares are held by the family holding company that also holds the controlling shares of VW AG. At the moment, the same person is also CEO of both companies.

So what happens if you have 22% of a company that owns 55% of a joint venture? I don't _think _ you actually get to take a decision on 22% of the 55%, do you? The 22% company's board would probably get together, decide on a single vote for the 55% percentage and your ownership would not be useful in decision making?

> Not sure if enough people have Rimac on the radar.

I've been aware of them for a while but they are largely irrelevant for most people. The technology is interesting and styling is cool but they were something like 1.9m euro.

For now Rimac has been most successful in finding investors. His game is similar to Musk in that he's very capable of swimming in the big money ocean, but the results are slim. This year they received 180 million from the state to develop autonomous taxis. It's not just state money, he's received private investment too, but it's all a bit fishy. Big claims and announcements, but a very slow rollout.

They have been building and selling automotive electric systems for posche/audi/ferrari and others for years, and have released their own super car. What are you talking about?

They've built a handful of ultra-expensive boutique automobiles. This is not nothing, but it's operating on a different scale than a company like Porsche. They have achieved world beating performance...on vehicles that are basically lab specimens that will hardly ever be driven.

Yeah, the guy said back in 2021 that they'll make lvl 5 (LOL) autonomous robo taxis by the end of 2023.

And the government bought the techbro investor-speak and gave an already shady company 180m € of taxpayers money.

https://www-telegram-hr.translate.goog/vijesti/telegram-dozn...

They're the Croatian equivalent of German "hidden champions" - insanely good technology, possibly world-leading positions in their niche, but very unknown outside of said niche.

Rimac cars aren't something your average neighbor will ever drive, probably not even your average non-F500 CEO. But, similar to F1 racing cars and early-days Tesla, they push right at the edges of technology, which will eventually trickle down to ordinary cars.

In Croatia they are also known of being shady company with shady government deals, shady beginnings (money laundering) and burning money on projects that aren't feasible at all. But they look good on paper and government lovea to use them as a PR and "successful Croatian project"

Thanks. The amazing campus was a bit too beautiful to be true for a growing car company.

This is fascinating to me.

What fraction of companies in Croatia are "money laundering"? Because if the fact is that most companies start out this way, and we end up with Rimac due to a magical grouping of people, then their success is hardly ignoble at all; it should be celebrated, supported, and leveraged into a moderate size organization. In that sense, they were successful despite their beginnings, which would normally be structured to not foster innovation. So they should be celebrated by Croatia and by Croatians since it is an example of what companies can be, and not so much an example of how companies can start since we already know the probability of success with that method is very low.

Rimac also did the battery system for "Koenigsegg Regera – the world’s most powerful production car" https://www.rimac-automobili.com/media/press-releases/rimac-...

In what sense is a Koenigsegg a production car?

It can be purchased by a consumer and driven on public roads.

In what sense is it not one?

It's hand built. And it would not surprise me if every one of them is different.

I can buy a Caterham-7 and drive it on public roads too. That doesn't make it a production car even though it is probably much more standardised than the Koenigsegg.

In the sense that there were 85 of them built on the same production line and sold to actual customers.

Usually this means that it is a street legal car that is/was sold to the public.

In modern times this also means that it went through crash testing so modern production cars have much larger production runs to offset the costs of that or are very very expensive (millions). For example Rimac said they are planning to make 150 Neveras but they also had to make multiple rolling bodies and a couple fully built cars (at minimum one for EU and one for US if you want to enter both markets) for crash testing on top of any development cars.

But strictly speaking there is no one definition but instead each field/industry defines it in its own way. Motorsports has its own (multiple actually depending on the series/governing body), car manufacturers have their own, record keeping organisations have their own (guinness book of records), etc

I feel like the minimum standard is too low, tbh. Building dozens of them and selling them seems like enough.

There needs to be another standard that is something like... built at least 10k of them. Because less than that is still likely hand-built and it doesn't really mean much that it was "production".

"Mass Market Production"?

In the sense that it is "homologated" and has had to pass all of the same tests as a car that will sell in the hundreds of thousands, or even millions of units.

https://www.tuvsud.com/en/industries/mobility-and-automotive...

> Homologation is the granting of approval by an official authority.

For anyone else to whom this word is news and didn't understand that the "and" meant "which can be translated to English as"

This word is common in latin based languages.

English is not latin-based but germanic afaik

The 60% of the vocabulary that came via French might have something to say about that.

Mate Rimac was also functionally placed in charge of Bugatti by Rimac and Bugatti becoming a joint venture.

Moreover, Porsche has invested heavily into the brand and currently has between a 45 - 58% stake in it following the formation of the Bugatti-Rimac JV. I'm giving a range for Porsche's ownership because there are multiple conflicting sources.

45% is the direct ownership, but Porsche also has indirect ownership on top.

From the Rimac press releases we have two different states with accompanying charts.

2021-07-05 - Porsche owns 45% of the joint venture directly, plus 24% of Rimac for an effective total of 58.2%. [1]

2021-11-02 - Porsche owns 45% of the joint venture directly, plus 22% of Rimac for an effective total of 57.1%. [2]

[1] https://www.rimac-automobili.com/media/press-releases/rimac-...

[2] https://www.rimac-automobili.com/media/press-releases/rimac-...

Im quite bothered by the “standard” one foot roll-out that car manufacturers advertise these day with their 0-60 times. It’s not 0-60 with it when there’s one foot roll out so it just feels unrealistic. And besides the difference in time measured between actual 0-60 and the one with the one foot roll-out is huge.

Thanks! I was not aware of it. It appears pervasive and ubiquitous, but completely slipped my radar. I'm barely a "hobbyist occasional track day racer" but it at least partially explains why I or anybody I know cannot come close to replicating the numbers I see (besides "they have much better drivers", of course:). First foot is HUGE, especially amongst different drivetrains (FWD/RWS/AWD, manual/DCT/AT/CVT), which handle the launch completely differently!

Without a rollout most 0-60 times would be useless. You make things like the surface the car is on matter way more than they should.

What's more of a problem is the opposite: if you increase the rollout so that the car's ECUs think it's being driven normally, and then mash the pedal to see how fast it goes, you get dramatically slower numbers: https://www.roadandtrack.com/car-culture/buying-maintenance/...

That reflects how fast a car will feel in daily driving as opposed to a drag strip

Thx; hilariously, I own both the Honda Odyssey and the Subaru WRX example cars from the article :)

>>"It’s that 5-to-60—the first number you should look at, and when the light turns green, the only one that matters. "

I think 0-60 from stop, 0-60 with foot rollout, and rolling start 5-60 are all valid and useful metrics, when explicitly indicated as such. While I've never participated in a public-street red-light drag race, I don't think 5-60 is a valid metric for it, precisely for the reasons it mentions - the 5-60 eliminates a lot of drivetrain and surface components which would impact the launch; and more directly compares just the pure power. My Subaru WRX has less power, but has AWD and grippy tires, will leave my 1-wheel-drive heavily-traction-controlled, all-season tires Minivan in the dust at launch (but things may become more equal from the rolling start or even on highway overtakes).

> more directly compares just the pure power

That's closer to what most people racing in a straight line want to measure as opposed to who can hook faster: hence drag racing using a two stage starting position

No metric will every reflect racing from a dig on a public street between the random surfaces, heat soak (it takes one or two pulls for most cars to start pulling timing stock), tires, etc... and I don't think any publication is really interested in satisfying that comparison.

I don't disagree. To your and authors point I have no interest in wrecking my daily driver transmission. I was just amused at the author spending 3 paragraphs explaining confounding factors that affect from-0 performance, and then offered the canonical from-0 situation, a red light, as suitable for 5-60 metric :-).

Anyhoo. Good read nonetheless :-)

>What's more of a problem is the opposite: if you increase the rollout so that the car's ECUs think it's being driven normally, and then mash the pedal to see how fast it goes, you get dramatically slower numbers

No, that's a good thing, maybe that would force them to not fuck up their throttle response!

I think that particular quirk is just car companies trying to make slow speed driving easier by filtering inputs more.

>Im quite bothered by the “standard” one foot roll-out that car manufacturers advertise these day with their 0-60 times. It’s not 0-60 with it when there’s one foot roll out so it just feels unrealistic. And besides the difference in time measured between actual 0-60 and the one with the one foot roll-out is huge.

Dumping clutch is also unrealistic to anything done daily. Or them most likely doing that test in absolute best conditions possible.

Personally I'd like more 5-60 test results so no traction control/clutch dumping or other tricks to shave fractions.

It's a pretty big claim, 280kWh/m^2. That's 4x the density of proposed iron-air batteries at utility scale (100h storage x 3MW/acre = 74kWh/m^2). However, I don't know if areal energy density is relevant to the market.

This battery is for a car, not grid storage (or at least this is my read)

I don't think this is for a car. There's a line in the article, "Details have been eagerly anticipated since the company announced it was going into stationary energy storage back in May".

Looks like it’s designed to charge EV’s at 400kW. Perfect for racetracks which may need insane charging rates but aren’t full time charging stations. 5 super cars at 400kW each would normally take an insane grid connection, but you can trickle charge these with solar and or vastly smaller grid connections.

It isn’t just race cars, this approximate power level is widely available in consumer cars.

Hyundai Ioniq-5 and other vehicles can utilize the increasingly common 350 kW connectors — including the Lucid Air, the Porsche Taycan, the Audi E-Tron GT, and the GMC Hummer EV. [0]

This enables them to deploy the fastest DC charging in places where the grid cannot support it. It’s a big deal.

[0] https://www.greencarreports.com/news/1135771_hyundai-ioniq-5...

You can use the battery on top of the existing grid connection. So if you can charge the 790kWh battery pack on a 50kW grid connection you can charge at 400kW + 50kW = 450kW which seems overkill if you only need 350kW.

It might also be designed to support multiple EV’s as most cars can only stay at 350kW for a relatively short period.

790kWh for a car? Now that should be sufficient range for everyone. But seriously, this isn't for cars.

>> the battery energy storage system (BESS) division of EV supercar company Rima <<

ESS == stationary Battery Energy Storage System. So this would be comparable to Tesal's Megapack business unit. I mean 790kW in an EV? That's must be for a rocket.

It is relevant, some areas will need higher density and others will need less density.

For bulk grid storage, batteries will likely be buried and not very dense.

For local spot backup for buildings and urban areas you absolutely want the highest power per cubic unit.

It seems to be in line with regular LiFePO4 batteries. Wikipedia is claiming 325 Wh/L^2 for Lithium iron phosphate batteries. This seems to be more about the inverter technology, which I guess is what is different here compared to a bunch of LiFePO4 cells in a box. There's a quote in the article about the inverters allowing them to get the "best footprint in the industry", so I'm guessing they managed to get the inverters down to about 45 litres in volume (or less).

How does per-square-meter work out to be the relevant unit for energy density? Where land is expensive wouldn't you just stack things up?

Those batteries would probably be installed in buildings, and you probably wouldn't stack them, so the relevant dimension is area required per kWh.

So am I reading this right, they are getting to 92% efficiency by knowing that their battery cells are insanely good low-pass filters that they can throw anything at?

No, they got there by designing insanely efficient integrated AC battery chargers and inverters at incredible power levels, which it seems no one has done before.

Otherwise all one would have to do is get the batteries.

Unless I've missed something, the primary factor for grid level storage is cost. You can make a super high tech, super efficient system, super dense, and none of that matter if it costs a lot (bar maybe some edge cases).

There's no mention of the actual electrical architecture. In the rendering, I only see modules and no DC/DC converters or inverters separate from the modules. Some of the competition uses module level inverters so maybe that's their approach as well? It's hard to tell if the 3 large conductor looking objects going down the font of the module faces (or rear?) are 3 phase AC or 2 x DC cabling + 1 comms or a fire suppressant line.

It would also be good to know what depth of discharge nets them 12,000 cycles (edit: looks like 95%)

No mention of costs. 12,000 cycles may be good, it may not be - how much are you paying per kwH per cycle in terms of replacement costs?

Reply on news.ycombinator.com