I absolutely will listen to this (and I'm sure learn a lot!), and also find it very funny and Volts-ian that this very wonky conversation dropped the morning after the election, it couldn't possibly be less aligned with the news cycle. I mean that as a compliment!
Ha we'll happily take that compliment. Counter-programming is a proud Volts tradition. Rest assured we have not entirely ignored the news though: David's election episode lands Monday, with what the results mean for clean energy and climate. Enjoy the wonk in the meantime!
There seems to be no real functional advantage to having smart SST transformers on neighborhood poles and the cost would be 10X more than Iron and Copper over the 50+ years that traditional transformers have achieved.
Grid tied rooftop solar invertors are already mandated to regulate their AC output and will automatically curtail. Smart thermostats and smart EV chargers are already the purview of the smart electrical meters that we already have and the smart SST will have no granular control advantage or functionality over these individual devices.
Few inverters match the long term reliability of basic dumb transformers. Inverters have 1000+ more failure modes. I'm not saying reliability is impossible, but it's rare to see a 25yr warranty on an inverter (usually 10yr), whereas transformers have a 25-35yr average, proven lifespan, with no need for firmware updates and associated software and control chip maintenance costs.
The iron-and-copper side of that comparison has gotten a lot pricier. The Labor Department's producer price index for transformer makers is up about 86% since January 2020, against about 31% for consumer prices, and utilities recover what they pay for that gear through the rates on our bills.
Replacing millions of analog iron-and-copper transformers with silicon-driven, solid-state modules is the strict thermodynamic prerequisite for building a truly democratic energy grid. We cannot wire a decentralized network of rooftop solar arrays, local battery banks, and electric transit systems across three million square miles using century-old magnetic coils that simply bleed heat and voltage. We need bidirectional power electronics capable of switching 800 volts of direct current at the microsecond level to seamlessly coordinate gigawatts of localized generation.
Right now, the manufacturing of these digital transformers is being bankrolled by the massive capital behind private server farms—corporations demanding hundreds of megawatts to monopolize computation and enclose our digital civil liberties. That hardware must ultimately belong to the public commons. A decentralized, solid-state grid allows local worker cooperatives and neighborhoods to control their own clean electricity. It breaks the monopoly of centralized utilities that still burn coal and gas to subsidize heavy polluters, including the massive industrial feedlots that take more than they give while caging animals that feel pain and fear.
We must champion this clean engineering to phase out the fossil fuels that wreck the physical climate, freeing up vast acreages to restore the forests and soil that soak up carbon. If we mandate open-source firmware for these 5-kilowatt modules and keep the physical infrastructure out of private hands, we can wire an energy system that serves universal human rights rather than concentrated corporate wealth.
I'm with ya - but there are big IFs in there. IF the firmware is open source. IF communities can advocate for their local needs. Since these solid state xformers are expensive I think the wealthy will invest in and own them, and exploit their data, to increase their control over the masses, not the other way around.
Spot on, @Caleb Rust. You are pointing straight at the central tension: whoever owns the physical switching hardware dictates the energy flow and extracts the local data. But we shouldn't let the currently high price of silicon carbide transistors convince us these modules will remain exclusive luxury assets.
Just like solar photovoltaics and lithium-iron-phosphate cells, the manufacturing cost of solid-state power electronics will relentlessly crash as production scales. Moving terawatt-hours of clean electricity physically requires deploying millions of these nodes. The immediate threat isn't that the wealthy will simply hoard them in private compounds; it’s that investor-owned utility monopolies will install them across our public rights-of-way, locking the firmware behind proprietary algorithms to surveil our usage and throttle local cooperative generation.
That is exactly why treating the grid as a democratized public commons is non-negotiable. If municipalities and worker cooperatives own their local microgrids, the data stays private and the infrastructure serves human needs, not shareholder dividends. We have to demand open-source protocols today, before centralized monopolies turn a thermodynamic necessity into a digital cage.
This interview talks about grid-hardening vis-a-vis attacks on the software component, but does anyone know if solid-state transformers are also more vulnerable relative to current transformers to nuclear EMPs or solar radiation storms? I've heard that current transformers are already a weak point. Wondering if new tech would exacerbate or mitigate the issue.
It appears that it won’t make a huge difference. The vulnerability lies mainly in the really big transformers on the transmission grid and nobody is replacing those with solid state transformers any time soon.
Raghu Belur didn't explain this very well. He said: "So it takes AC in and puts AC out at different voltages and does it very, very well. The things that it doesn’t do is transform to different voltages. So for example, can I get AC in and generate DC out?"
What he meant is that the output voltage is constant and cannot be varied, and that the output is AC and can't be DC.
Solid-state transformers are the unglamorous hardware that decides how fast new load can connect. Do you think the first volume use is in data centres or on the distribution grid?
I’m curious about the proposed electrical architecture for utility scale solar.
In a utility scale solar plant I gather you have strings of panels wired up in series, (which adds the voltage up to something like 1000-1500 VDC) then inverters to convert from DC to AC, then transformers to convert to 33 kV (from memory) AC which then feeds into a substation to connect the the grid at 66Kv or more.
Are they proposing to get rid of the transformers in the substation?
Let's not underappreciate the magic that happens within a common transformer. That magic is called electromagnetic induction, discovered by Michael Faraday. IMO, he is the most influential person in history. Without electromagnetic induction there would be no motors and until PV, there would be no generation.
I’m curious to think what voltages a future home may be. Would we adopt something like a 400V or 800V DC to integrate solar, battery, EV charger, and heat pump, and then send 110/240VAC, or even 46VDC to the rest of the house. Any pro/con for going all the way to 800V or some lower voltage?
I think the future will look something like USB (or Enphase!). You plug in, gadgets talk about it and decide what they need to do, then do it. No significant potential will be on any connector contacts until a handshake is done.
I absolutely will listen to this (and I'm sure learn a lot!), and also find it very funny and Volts-ian that this very wonky conversation dropped the morning after the election, it couldn't possibly be less aligned with the news cycle. I mean that as a compliment!
Ha we'll happily take that compliment. Counter-programming is a proud Volts tradition. Rest assured we have not entirely ignored the news though: David's election episode lands Monday, with what the results mean for clean energy and climate. Enjoy the wonk in the meantime!
There seems to be no real functional advantage to having smart SST transformers on neighborhood poles and the cost would be 10X more than Iron and Copper over the 50+ years that traditional transformers have achieved.
Grid tied rooftop solar invertors are already mandated to regulate their AC output and will automatically curtail. Smart thermostats and smart EV chargers are already the purview of the smart electrical meters that we already have and the smart SST will have no granular control advantage or functionality over these individual devices.
Few inverters match the long term reliability of basic dumb transformers. Inverters have 1000+ more failure modes. I'm not saying reliability is impossible, but it's rare to see a 25yr warranty on an inverter (usually 10yr), whereas transformers have a 25-35yr average, proven lifespan, with no need for firmware updates and associated software and control chip maintenance costs.
The iron-and-copper side of that comparison has gotten a lot pricier. The Labor Department's producer price index for transformer makers is up about 86% since January 2020, against about 31% for consumer prices, and utilities recover what they pay for that gear through the rates on our bills.
Replacing millions of analog iron-and-copper transformers with silicon-driven, solid-state modules is the strict thermodynamic prerequisite for building a truly democratic energy grid. We cannot wire a decentralized network of rooftop solar arrays, local battery banks, and electric transit systems across three million square miles using century-old magnetic coils that simply bleed heat and voltage. We need bidirectional power electronics capable of switching 800 volts of direct current at the microsecond level to seamlessly coordinate gigawatts of localized generation.
Right now, the manufacturing of these digital transformers is being bankrolled by the massive capital behind private server farms—corporations demanding hundreds of megawatts to monopolize computation and enclose our digital civil liberties. That hardware must ultimately belong to the public commons. A decentralized, solid-state grid allows local worker cooperatives and neighborhoods to control their own clean electricity. It breaks the monopoly of centralized utilities that still burn coal and gas to subsidize heavy polluters, including the massive industrial feedlots that take more than they give while caging animals that feel pain and fear.
We must champion this clean engineering to phase out the fossil fuels that wreck the physical climate, freeing up vast acreages to restore the forests and soil that soak up carbon. If we mandate open-source firmware for these 5-kilowatt modules and keep the physical infrastructure out of private hands, we can wire an energy system that serves universal human rights rather than concentrated corporate wealth.
I'm with ya - but there are big IFs in there. IF the firmware is open source. IF communities can advocate for their local needs. Since these solid state xformers are expensive I think the wealthy will invest in and own them, and exploit their data, to increase their control over the masses, not the other way around.
Spot on, @Caleb Rust. You are pointing straight at the central tension: whoever owns the physical switching hardware dictates the energy flow and extracts the local data. But we shouldn't let the currently high price of silicon carbide transistors convince us these modules will remain exclusive luxury assets.
Just like solar photovoltaics and lithium-iron-phosphate cells, the manufacturing cost of solid-state power electronics will relentlessly crash as production scales. Moving terawatt-hours of clean electricity physically requires deploying millions of these nodes. The immediate threat isn't that the wealthy will simply hoard them in private compounds; it’s that investor-owned utility monopolies will install them across our public rights-of-way, locking the firmware behind proprietary algorithms to surveil our usage and throttle local cooperative generation.
That is exactly why treating the grid as a democratized public commons is non-negotiable. If municipalities and worker cooperatives own their local microgrids, the data stays private and the infrastructure serves human needs, not shareholder dividends. We have to demand open-source protocols today, before centralized monopolies turn a thermodynamic necessity into a digital cage.
This interview talks about grid-hardening vis-a-vis attacks on the software component, but does anyone know if solid-state transformers are also more vulnerable relative to current transformers to nuclear EMPs or solar radiation storms? I've heard that current transformers are already a weak point. Wondering if new tech would exacerbate or mitigate the issue.
It appears that it won’t make a huge difference. The vulnerability lies mainly in the really big transformers on the transmission grid and nobody is replacing those with solid state transformers any time soon.
Raghu Belur didn't explain this very well. He said: "So it takes AC in and puts AC out at different voltages and does it very, very well. The things that it doesn’t do is transform to different voltages. So for example, can I get AC in and generate DC out?"
What he meant is that the output voltage is constant and cannot be varied, and that the output is AC and can't be DC.
Raghu wants a transformer that can multitask.
Solid-state transformers are the unglamorous hardware that decides how fast new load can connect. Do you think the first volume use is in data centres or on the distribution grid?
I’m curious about the proposed electrical architecture for utility scale solar.
In a utility scale solar plant I gather you have strings of panels wired up in series, (which adds the voltage up to something like 1000-1500 VDC) then inverters to convert from DC to AC, then transformers to convert to 33 kV (from memory) AC which then feeds into a substation to connect the the grid at 66Kv or more.
Are they proposing to get rid of the transformers in the substation?
Let's not underappreciate the magic that happens within a common transformer. That magic is called electromagnetic induction, discovered by Michael Faraday. IMO, he is the most influential person in history. Without electromagnetic induction there would be no motors and until PV, there would be no generation.
I’m curious to think what voltages a future home may be. Would we adopt something like a 400V or 800V DC to integrate solar, battery, EV charger, and heat pump, and then send 110/240VAC, or even 46VDC to the rest of the house. Any pro/con for going all the way to 800V or some lower voltage?
I think the future will look something like USB (or Enphase!). You plug in, gadgets talk about it and decide what they need to do, then do it. No significant potential will be on any connector contacts until a handshake is done.