📖Signed copies of my book, Engineering in Plain Sight, are back on sale! store.practical.engineering/ 🌌Get Nebula using my link for 40% off an annual subscription: go.nebula.tv/Practical-Engineering
Oh snap ... I've never thought of distributed generation solar increasing rocof. It's a lose lose situation for grid tied solar. The next Gen nat gas turbines (GE and Siemens) seem really cool, and a good mitigation for this issue
In early 2025, the Baltic states in North-Eastern Europe are set to disconnect their grid from the Russian power grid, and connect up with European power grid instead. This might sound like it's simply some sort of resynchronization task, however the reality is much more complicated, requiring the Baltic states to build out a fair bit of extra infrastructure and take on the task of maintaining grid frequency where before, Russia would have been doing so. This might make for an interesting video.
I live at a junction between two power companies' service areas, and they have the ability, in an emergency, to connect one to the other, but there is the catch that there has to be a transformer, there, because the operating voltages don't match - and there's no synchronization, and there's a meter to measure the power that crosses junction.
@@kenbrown2808 Brazil has the same issue, along with Japan, and built HVDC systems that are both frequency changing, and bidirectional, so that power can flow either way, along with changing the frequency of that power to suit the local grids as well.
I work as a lineman in northern EU, and an interesting effect we've noticed with solar is that sometimes the inverters won't disconnect from the grid during a power outage. We've ran into this when doing service work on pad-mount transformers, when we open the switch to de-energize the transformer supplying a neighborhood (in our case an average of 150 households/transformer) with a lot of solar, on rare occasions it will remain energized, backfed from the households solar production. This happens because in the right conditions the solar panels will produce the same amount of power as the households consume, so there's no current flow through the transformer. So, when we open the switch - as far as the inverters are concerned nothing changes, so they remain on and backfeed the grid. It usually doesn't last very long since when the balance shifts they'll disconnect but you could still get a nasty surprise if you're not careful.
Ah now this is technically illegal, all devices that back feed to the grid must have anti-islanding devices fitted. Well in France and the UK. When I worked for RTe (french transmission operator) there was talk of taking some people to court etc about this issue. The same goes for houses fitted with a gen, it must separate from the grid when the grid goes down.
Woah, that's an interesting fault. I've never heard / thought about this exact configuration, but yeah, they would stay on as long as that configuration remains stable...
@@matthewmaxwell-burton4549 Yeah we have the same here but the issue is them not being able to tell if the grid went down, as long as the flow stays balanced at least
I learned today that new installation of a grid-interactive inverter must: have ability to throttle the power and also be able to shutdown the Inverter power and control the scheduling of when the power is used/delivered and the quantity on command from the utility. I imagine the direct command & control will tend to resolve this accidental self-supporting meta-stable microgrid situation.
This was interesting. I was a nuclear trained submarine officer and served on an improved LA class sub for 3 years. Going through the nuclear training pipeline, you learn a lot of this stuff. In fact, the crude demonstration in the video of a drill turning an AC generator is something we had two: the "ships service motor generator" (SSMG -- everything needs an abbreviation), which were approximately the size of a midsize vehicle. They converted power back and forth from the AC to DC bus, where the DC bus was powered by **VERY LARGE** lead-acid batteries in the bottom of the sub. Very interesting stuff. The control room ("Maneuvering") had the panel for managing the electrical buses, and the operators manually would synchronize various bus frequencies ("slow, in the fast direction" for the incoming bus, and throw the switch for the breaker at approximately the 10/11 o'clock position to shut the breaker at 12). Depended on the operator -- not automatic. I believe newer subs use a solid state inverter system and not the SSMGs now. Anyone really interested in hands on, great education in this sort of stuff, either as enlisted or officer, join the Navy and go to Nuclear Power School and get to a ship. I wasn't an engineer in college, but you more or less become one working the engine room of a nuclear submarine.
I taught NNPS when it was in Orlando. After that, I designed and built the electric plant trainers based on the S5W (Bluefish, et. al.) plant used in NFAS back in the late 80's. Modeling the SSMG was one of the more challenging things I've ever done as an engineer. I used capacitor charging curves to emulate the starting surges that occur when the resistor sequence is pulled as the rotor speed comes up. The ships battery was a 3 D-Cell actual battery pack, that was literally charged and discharged from the op amps of the model. Overall that project was the best engineering task I ever undertook, and got a Westinghouse Signature award for the work. But they wouldn't promote me ahead of "years in grade" ( just like the military!) so I quit. And BTW both Bettis and KAPL were absolutely bursting with "engineers" that couldn't design a resistor divider.....but they wrote beautiful reports to be sent along to Navsea-08 to impress them with how much was being accomplished. 😮
@@jeffferguson4632 Oh, that's neat! I was one of the last classes through Orlando NNPS in 1997. Your comment about the resistors in the SSMG startup is memorable. The DC startup on those machines caused the biggest transients on the electrical panel -- always something to watch when that happened. Plus, I recall a teachable moment about good engineering practice when we restarted an SSMG (following routine maintenance) while in an unusual electrical lineup. That's quite a thing to have been involved in the engineering of. They worked well from my perspective.
I was involved in the early design of the Moored Training Ship setup in Charleston....when the simulator was located in the trailer. You wouldn't believe the arguments we had about putting a "head" in the trailer. If you went to MTS Charleston and bitched everytime you had to go back to the building to take a leak....well...I tried.
I am a test engineer for high power frequency converters and be build our "VSD"'s not only for grid injectors like wind turbines (exactly like in the video), but also for applications where you would not suspect them. Is is for example sometimes easier to let a water turbine run at a non-constant speed and run all that power through a VSD, which constantly synchronizes to the grid, than try to maintain a constant speed with the turbine. The are also used as a soft starter for these very big motors and sometimes, they can even reverse the flow of energy to pump water back up into a lake for example. We are talking from 500kW to 100MW here. Great video!
I worked at a factory that ran 24/7 and consumed megawatts of power. We were looking for backup power solutions. The problem was the few seconds backup generators take to start. There were "batteries" that consisted of carbon fiber flywheels in a vacuum spinning at 500,000 RPM. They could deliver 100% output (multiple megawatts) within 10 cycles, and for up to 10 seconds - enough time to start a generator kept in warm stand-by. Perhaps technology like this is what we need to stabilize the grid.
The UK is spending billions in building flywheels to pick up the janky electricity from solar and wind and add a modicum of control to it. Or we could just use nuclear 🤷♂️
There are a lot of companies looking into grid scale flywheel energy storage. It's a valuable option and one I could stand to learn more about. But I think the biggest problem with any **grid scale** energy storage solution is cost. I'd love to see a practical engineering deep dive into grid scale storage and fly wheels.
I understand that flywheels are already used by various utilities for this sort of purpose, providing inertia at a moment's notice. Wikipedia at least says that there are grid-connected flywheels in New York, Germany, and Ontario. Would be nice to see more about how these sort of utilities work!
Yeah, I am hardly an expert, I just remembered the Capex research I did for that project a decade ago. The thing with flywheels is stored energy goes up linearly with mass, but squared with rotational speed. So a 1 pound carbon fiber disk spinning at half a million RPM can potentially store more energy than a multi ton wheel spinning standard generator speeds like 1800 RPM. Chances are people in the industry know much more about the details than I do, but occasionally good ideas slip through the cracks.
Flywheel storage solutions are interesting. I work on mission critical power systems, and most of the time a battery based UPS is used. For certain loads, especially at a factory where there are huge motors flywheel based solutions can be better
My issue with solar: all we are doing is tearing down forests to build them… lets destroy more habitat so we can build solar power, yay!… looking at you australia
Just out of curiosity do batteries help "clean up" the power from solar? I've heard one issue is the power they put out can sometimes be "dirty" and cause issues. I'm thinking of solar at home but I've heard those systems can sometimes cause issues for peoples neighbors. I'm not sure with the inverters that people put in along with a battery backup can sorta smooth everything out so the power being output is clean. From this video it seems like the inverter might actually be the issue. As a side note would flywheel storage being added to larger renewable systems help with this or is that still needing to go through the inverter? The flywheel system might not be actually used for storage but it could just smooth out the output? Seems like wind might not have the same issue though since the turbine itself almost acts like a flywheel.
@@pin65371 Batteries basically use the same inverters as PV. Either literally the same inverters (if DC coupled) or models that were likely derived from PV inverters (if AC coupled). Either way, I disagree that the output from PV facilities is "dirty". As the video mentions, it can output a perfect 60hz using electronics alone. They can also be called upon to inject or absorb VARs to help strengthen the grid. The real issue is that renewables are intermittent which can make load balancing challenging for grid operators. Batteries definitely help with this, as they can help firm or smooth the output of the plant to avoid fluctuations. I've never heard of flywheels being used alongside PV, but some companies are looking at alternatives to batteries, such as hydrogen electrolyzers to "store" energy as hydrogen.
@goldenhate6649 That is far from what we commonly do in the US at least for my EPC. Additionally when there are protected areas or wildlife we often have to have a dedicated biologist on site. SWPP is also huge to ensure runoff does not cause problems for thr surrounding area.
the final project for one of my classes was to design a solar installation for a factory, we had to calculate power usage from power bills, design a capacitor bank for power factor correction, find out how many panels, in what arrangement, and at what angle and distance from each other they had to be installed, pick an MPPT and inverter, and design a battery bank. i couldn't finish in time, it was a nightmare, but at least it taught me that people way underestimate how much work goes into installing solar
Anyone who has tried to design their own solar panels and inverter setup for their home knows that its a complex process that has you reading datasheets for the panels and the manuals for the inverter!
@@tlangdon12I ordered a kit online in 2017, 20 Astro Energy 310W modules, DC optimizers, and a Solaredge SE6000A-US-U inverter, Iron Ridge XR100 racking, and connected it all. I got a Bosh D-Tect 150 to help me hit the rafters through the roof with the 5 inch stainless lag screws holding the flash feet. Making 4.47kw right now. It's all tested by UL. There was a little paperwork to get permits, and pass inspection. Your backfeed breaker has to go in the bottom because the solar adds to the amount of power you can pull out of the pannel, and you don't want it to add that to grid power and exceed the capacity of the bus bar. If you put it in a subpannel, same thing, and it needs to be in the bottom of the main pannel. It's not a beginner project, but totally doable if you are motivated.
Complexity varies. I just pluged microinverter to grid and screwd panels to frame and done. Nothing complex on simple house system. Not a single thought about power factor correction. Its how complex do you want or need to make it.
I learned all about MPPT while shopping for solar charge controllers. No tutorial I found actually showed the same panel in different conditions, they just said that the MPP changes. Your test results were nice to see, and helped with understanding. Thanks!
The Odessa event was (edit: contributed to) by a particular phenomenon called Cessation, which is not an unexpected behavior of nebulous algorithms. Designers put it in intentionally to protect the hardware without violating the grid rules at the time, which say when a disturbance happens, you must keep your output breaker closed to help the system (low voltage ride-through.) But there were no requirements that your unit continue to output (real) power. Solar units therefore would drop their power output to nearly 0 to protect the hardware from harmonics generated by the panels trying to push power into a disturbed grid. There was also an automatic timeout on this cessation where the panel would not ramp up until the disturbance had passed for some time. This was often baked into the hardware and many farm owners didn't even know the behavior existed. 6:30 this explanation would have benefitted a lot from showing a raw PWM signal, then showing a capacitor low-pass filter wipe across it to smooth it out
@@rowanjones3476 eh, it was more a case of hardware designers not being aware of WHY particular requirements exist. They met the letter of the law while violating the spirit of it.
@@marshallc6215 As a hardware designer, this kind of thing drives me nuts. We're always having to go back to product managers and ask WHY they want things on our boards. But I understand not everyone feels like they have that ability, or perhaps are jaded after being told "just build it to the spec and quit bugging me!"
The key thing here is that the inverter has to push real power. If all you needed was a "sine wave" then you could produce one with a DAC and a simple capacitor-based filter. The problem is that a DAC is a resistive device and can't actually push any real power (it would melt into a pile of sludge in milliseconds if it tried), so the sine wave is useless for inverter purposes. For the same reason, using e.g. a transistor type of circuit to amplify such a loe-poert sine wave won't work because the transistor is a diode-like device and the voltage difference is dissipated as heat. The transistor can't actually amplify a DC input and produce a ton of power on the output that looks like a sine wave either. It would instantly melt. Same thing for a PWM signal. The PWM signal itself can drive FET gates, and the FETs are what switch the real power from the DC or rectified source. But a capacitor on that output won't produce a sine wave, it will just melt (more or less instantly) because it is essentially short-circuiting the FETs. Again its an issue of the power flow. A capacitor that isn't doing any work won't melt, but in this case the capacitor is doing a massive amount of work and the FETs are basically pushing incredible currents into and out of it because it kinda looks like a short to the FETs (hundreds of amps.... 1000VA for each real 1W of power). So the only way to really produce a real sine wave with the desired amount of power is with magnetic components... inductors and transformers. The PWM can drive the FETs, the FETs run the real power through an inductor to convert the medium-frequency PWM signal into a voltage, and that runs through a transformer to produce the desired output AC voltage. The inductor and the transformer both filter what is essentially a huge HF mess of square and sawtooth waves into something close to a pure sine wave and a single capacitor on the output gets rid of remaining HF artifacts. -Matt
@@junkerzn7312 Ehhh, a modern cheap grid-tie inverter for areas that are 3-phase would just be a triplet of half bridges made of SiC FETs (rated to 3x the phase-to-neutral RMS, measured across the entire half-bridge), with an LCL low-pass between the switching node and the grid-side phase wire, with the PV strings MPPT boosted up to said DC level (e.g., in central Europe with about 220~250 V, we'd be at about 750 V for the DC bus). At larger powers you will want more of the SiC switches in the 3-phase converter section, because the extra cost to control them all separately will be worth the savings in the LCL low-pass filter that theoretically losslessly eats up the PWM's higher-frequency parts only letting the low-frequency smoothed sine through. But you need around 50-100 kW (well, I guess, kVA) before that should be worth the complexity.
power electronics engineer here: Good video. I do have a few additional thoughts, though: 1. You said a home PV system cannot provide power during a blackout. At the end of the video you talk about grid forming inverters being able to support e.g. a backstart or island grids. The good news is: Those systems do exist for home PV applications. They are usually more expensive, but can supply power during a blackout. (There are different types being able to do different things exactly - I'm not going into details.) 2. One problem not addressed in the video (or maybe I missed it?) is the fact that mechanical generators can provide high short circuit currents that will trip breakers (not the ones in your house, the large ones being part of the power grid). Since fault currents have always been high in the past (something like 10x the maximum normal current), circuit breakers were designed to not only handle them, but also require them for fault detection. But: Inverters don't deliver such high surge currents, even if they stay active during the fault event. They generally limit their own output current. While a mechanical generator can survive a current surge into the grid, a - for example - PV inverter couldn't (hence the current limiting). This is different from fault ride through (FRT) and is an additional problem. IMHO the solution to this is more modern protective equipment that will detect faults differently and trip not only during high current surges. 3. What you're showing at 6:06 is NOT pulse width modulation (PWM)! This is (kind of) what a multi-level inverter's output voltage could look like. A PWM still only has two voltages: plus and minus dc voltage (in the case of a two level inverter, that is), but switches between them rapidly, so that the resulting voltage resembles a sine-wave after filtering.
Grid operators usually don't want solar supplying power during a blackout. At best it complicates the process of reconnecting power, and at worst it makes cutting power for servicing very difficult.
The problem with demonstrating PWM is that the fundamental frequency of the modulator is usually off into the hundreds of kilohertz, but the frequency of the final output is in the tens of hertz. This makes is it impossible to show both parts at the same time, The levels of zoom are just too far apart. I think showing that off would have made the video too complicated for its intended audience. Also, he didn't say that what he was showing was PWM anyway. He said that it was multilevel output, although I'll grant you that describing PWM as being just like multilevel but with more switching doesn't really make any sense. And, the way that he describes it sounds a lot more like pulse density modulation, which is quite rare, as far as I can tell, but, in my opinion is easier to explain than PWM. PWM is just way more common, probably because it's a bit more simple in how the algorithm that drives it works.
I mean, if you know all this stuff you’re watching this mostly for entertainment purposes anyway. If you don’t know this stuff, that’s outside of the scope of this video. I like these videos because they give a broad overview of a subject.
1. He said "most" residential solar systems can't provide power during a blackout, which is true. You need a Tesla Powerwall, a Generac battery/inverter system, etc. to have both the storage and the grid-forming inverter to do that, and you need an automatic transfer switch to prevent you from energizing the grid during a blackout. 3. He specifically said this was from a "cheap" inverter (i.e. "modified" sine wave), not from a PWM inverter. There are multi-level PWM circuits. In fact, all 2-level 3-phase PWM inverters (which is most of those described in the video) are effectively 3-level because of the 3-phases. A 3-level 3-phase inverter is effectively 5-level.
@@gay4femboysNah, if you actually know how electronics work, especially modern small ones, you'd be blown away how much of a miracle that things work at all. Beyond walking in a knifes edge, you're walking on an edge literally hundred if atoms thin. So many insignificant things can make it all go haywire, our maths are amazing
@@Mcfunface that's ok for filling in the gaps right now while long duration storage is developed. But we have a lot of simple cycle capacity throughout the US so I'm skeptical of the need to build out a whole bunch more. Really what's needed more than new gas plants is better transmission between ISOs/RTOs and a simplified process for getting lines approved and built
Another old, now resolved problem with PV inverters was that in case of an over or underfrequency (I'm going with overfrequency for now), not just one or a couple inverters starting dropping their power but every single inverter connected to the grid. In the early days the power reductions level were hard set which led to situations where an overfrequency triggered thousand of inverters at once to drop their output, causing a sudden drop in the frequency, which then again led all the inverters to ramp back up, causing a fast over and underfrequency swinging, where the overfrequency swinging also led to wind turbines being shut down in the thousands, making the problem even worse
Power grid automation and protection are the industry I work in (R&D at Schweitzer Engineering Laboratories). This is one of the really pressing and interesting challenges which modern grid operators are dealing with. Wind and solar both add quite a few challenges when it comes to grid stability, but the economics are such that they are going to continue to become a larger piece of the puzzle moving forward, so we have to continue learning how to protect and regulate our evolving power grid. This was a very nicely produced video which strikes a good balance of talking about the challenges and reasons why power production still is going in this direction.
@@SeeNickView I think so. I love getting to work with the talented team of engineers I support, and really appreciate working for a company which has such strong values (and really tries to live them). I really appreciate how much emphasis there is on doing the best we can to always do right by our suppliers, employees, and customers. And getting to work on some really cool products and help solve fascinating puzzles is just icing on the cake.
Love your products.... The rtac software can use some work but the 351 series is my favorite relay, also really enjoy the 401 technology for fault detection and the new port servers.... Keep up the amazing work over there!
I work at a Public Utility District along the Columbia River in the Pacific Northwest. We are also a Balancing Authority for WECC. There are many systems that are always being monitored but load and it's effect on the frequency impacted by that demand is closely correlated and adjusted by adding or removing generation capability on the grid.
Another great video Grady. I've worked in the inverter industry for over 25 years and this video has taken all the complicated technology and simplified it to the point that a larger audience can gain knowledge. I'll be using this video as a supplement to the training material I've created. Well done sir!
Matt, Greg from Alaska here. Hope things are going well. Talked to Lones every few years :-) maybe he’s the only old OB tech left. Anyway, take care of yourself,
This helps me to appreciate what is going on inside my off grid solar system controller. Cooling fans spooling up and down in proportion to the amount of sun juice coming from the panels, amp meters on battery management units counting out the juice coming and going to keep me connected, comfortable, and conserving in an independent way. Super technology and we are gleaning the results of a massive effort. Thanks for being the one who explains the mega-systems under construction today.
Fun fact about the frequency. Microwave ovens usually use grid frequency to keep clock in time. That's why microwave ovens' clock never stays on time, at least very long. Where I live the clock ticks ever so slightly too fast, because our grid provides slightly over 50Hz power, but MWO still ticks on one second every 50 oscillations.
We have a 60hz grid and the operators actually average out the frequency by making sure there is 60hz*60sec*60min*24hours of cycles per day. Over time the number of cycles is about spot on, so our grid-based time keeping stays spot on over time, even if it varies a tiny bit throughout the day. Short of a power-outage throwing that all out the window, the only time I adjust the clocks is evil daylight-savings.
Yeah, the 50 Hz in Europe may vary a bit, but the frequency target is actually biased such that the summed cycle offset tends towards zero. There was an issue a couple years back when a dispute between Kosovo and Serbia resulted in underproduction and a prolonged lowered grid frequency. The noticeable effect of this was that frequency-tied clocks were running 6 minutes late. They regained this time over the following months though, as designed.
In my experience, there's 2 reasons a frequency-tied clock can get a permanent offset: 1) Your local area was islanded and the grid frequency was shifted to trip solar inverters. The change may be as much as 2%, resulting in clocks rapidly gaining time. 2) A simple blackout: No grid means no cycles.
@@AlexandervanGessel My microwave oven has been being running fast since we bought it over a year ago. It's not much, like 1-4 minutes a month, but never runs late. I'm not sure if the Kosovo-Serbian situation affected us, we are in Nordpool and AFAIK we use almost exclusively our own electricity, and import rest from Sweden. I can be wrong 'tho
@@TheSanpletext Well, I'm referring to the central Europe synchronous grid (which stretches from Morocco through continental Europe to Turkey). While Norway, Sweden and Finland are members of ENTSO-E, they have their own separate synchronous grid, which includes parts of Denmark. Maybe the frequency is less controlled there?
Frequency reserve response (FRR) is the holdback. Hydro units shine in this area given they have the largest rotating mass. When an event is triggered the MW setpoint will be released and the governor will respond by either picking up or dropping load (the event can be over or under frequency).
I work in power generation ( a good portion of my work is in the renewable sector). Solar farms present a unique challenge in the fact that there is no inertia behind the power production. A regular power plant ( gas or steam turbine) has an effect almost of a synchronous condenser ( a small power reserve). Inverter based generation doesn’t provide that. Which can be weird once you start getting into crazy demand days. Most inverters used on the grid are what’s called voltage source inverters, meaning that the dc voltage MUST always be higher than the inverter AC output voltage. This is to ensure the inverter can “pump” power out to the grid. From a dispatch standpoint , most solar farms are not dispatched, they produce what they produce to the grid. The ac voltage that gets exported to the grid is a relatively stable voltage regardless of the dc production, although the power flow will adjust with dc production. There is very little tolerances when it comes to grid tied systems. Most solar inverters and battery storage inverters ( again utility scale) are the exact same equipment. They can be configured for each operation. Almost all inverters from the major manufacturers, are capable of grid forming, which means they can start with no grid reference. These units can be paralleled to create a backbone to which thermal generation can actually sync to. Basically the more solar farms that pop up, the more battery storage systems will be needed. All utility scale inverters have some sort or primary, secondary, and reserve frequency response algorithms built in. They’ll just convert this mismatch into heat ( vars). Let me know if you have any questions. I can provide clear explanations on anything else. 😊
How big is the battery required for grid-forming? I'm guessing not by much, as it's essentially used as giant capacitor. Perhaps supercapacitor can be used instead?
@@hanifarroisimukhlis5989 they aren’t usually separate pieces of equipment. A battery storage facility has 2 electrical measurements. Power: expressed in megawatts, which is the amount of real time power that can be produced. Energy: expressed in MwH, or how long it can “generate” The black start BESS systems I’m currently working on are in the range of 3-5 megawatt systems, with 20MwH of capacity. So it can discharge at full capacity (5MW) for 4 hours. You need capacity and power for a black start to work, usually black starting the transmission grid takes time, even if successful the first attempt. You’re looking at a minimum of 3 hours to restoration depending on what caused the grid collapse
No grid tie inverter I have ever worked with works without "grid reference". Id est, without the 50hz reference point they cannot output. What century are you from?
I work in Scada for an EPC, those are good points. Would have liked him to dive a bit deeper. Seemed most of the video presented a challenge that none of our jobs usually face (the grid following). All of my jobs have PFR
Great job considering your world of Civil Engineering is quite a bit different wheelhouse than my world of Electrical Engineering. The depth that you dove into the EE aspects of power is very much appreciated by your entire audience, regardless of their profession.
tbh, modern engineering education (by 'engineering' here I mean application of physical sciences) should be changed in a way that it prepares students capable of understanding all forms of engineering as they proceed in their degree and could work in any one of them they choose, of course, they will undergo training as they start in the industry after finishing academia. They will have more clarity and higher intellect.
I learned a lot about the power grid from this video. The graphics were super helpful to conceptualize. It was also telling to see the choppy waveform of cheap AC inverters. Better than a square wave, but not by much!
Excellent coverage of a very technical subject. I worked as a protection engineer for 42 years. Originally when small wind machines were being added to the system it was common to have them trip off for any perturbation on thw system. But as wind and solar grew to be a significant percentage of generation we began requiring ride through ability. Of course the vendors all had their own designs and everything was proprietary and often didn't react as expected by us. Inverters also create issues for protection engineers as they don't produce fault current, they are current limited devices. That makes conventional relay protection unsuitable and more expensive high speed communication based schemes are needed.
IIRC fault-ride-through is actually designed to inject reactive power at possibly full kVA rating when a fault voltage pattern is detected, to trip the breakers without needing to have excessive real power on hand.
Another exceptional piece of work, Grady. I’m proud to be a patron of your channel. My Father-in-love was head load dispatcher for PECO back in the day (retired before Y2K) and I’ve shared this installment with him. I am sure he will enjoy it. Blessings, my friend. Keep up the great work!
It's even harder when power companies like PECO don't know how to: read solar schematics, inspect a customer's transformer size accurately, or keep to their own scheduled appointments. We've had panels sitting on a customer roof for 4 months because of the incompetence (maybe even greed) of the power companies. Meanwhile, this customer could have been 100% off the grid and feeding excess back into the grid which desperately needs the extra power!
Well, you have to look at it from the perspective of the power company. Being forced to purchase electricity from every customer with a system is not a good deal for them in any way. Even if you set aside the fact that having numerous producers whcih you have very little control over complicates operating your system and makes it more expensive, just the purchase alone (separate from the operational complications) is a raw deal. The power company is required to purchase the electricity at rates based on their retail rates. Here's the thing though, those retail rates are a combination of generation costs and the power company's transmission and distribution costs. Building, operating and maintaining the infrastructure that delivers the electricity to the customer's home is a significant cost. 50% is a common figure, but I'm sure that it varies depending on the situation. Regardless, the price you pay at your meter is partly the cost to generate that electricity, and partly the cost to get that electricity from the generator to your house. Requiring the power company to purchase electricity you generate at that same (or similar) price means that they are paying way more than the generation cost for essentially additional generation capacity, while ignoring the cost to them to distribute "your" power to other customers. So given that, it's understandable that power companies aren't thrilled to be required to enter into agreements that just cost them money, and understandably aren't leaping immediately to the opportunity to lose more money.
One of the better videos I have seen in a long time. You take a simple sounding project, like adding solar to the grid, and clearly explain many of the subtle problems that come up. I am going to use this video as a prime example of how simple projects can grow into complex problems. Thanks for all the hard work.
Many thanks to the people who have created this massive system, and who maintain it. I’m sitting comfortably in my air conditioned house watching RUclips videos on a Saturday morning thanks to all of you!!!
I found this super interesting! Usually you only hear about the problems with new technologies, especially when there's a powerful political lobby against them. So I was under the impression that renewables are an unqualified negative for grid stability and that we'd need flywheels or something if we get rid of too many turbine-based power plants. And I don't live in an anti-renewable bubble mind you! So thank you for teaching me that these new types of power plants aren't better or worse, they're _different_ and we need to learn how to integrate them well into our existing infrastructure while it is changing at a very fast pace. This seems to be a big challenge to engineers and regulators alike, but one that does not seem to be beyond their abilities. Well, not beyond that of the engineers at least.
Fantastic Video. Grady... you really should add a link to the black start video in the description for this one! Maybe even at the end credits. There is a strong connection here.
It's amazing how much smarter the grid is becoming all the time. To think - a hundred and fifty years ago all we wanted was electric lights that could withstand a lot of variability and now we have all this delicate machinery in our homes that could be thrown off by a few volts at the wrong time.
Depends, most DC items use a switch mode power supply, which are very forgiving. They're rated (usually) 100-240v 50-60hz, so anything near or within that range will be just fine. The most sensitive common item will be things like compressors and other motors
Actually more of the "smart" devices are problematic for the grid during brownouts, not the other way around. Switching power supplies will try to draw the same amount of power regardless of the input voltage. Normally if the grid sags, motors and basic resistive loads will draw less power and it can stabilize, but too many adaptive loads will pull ever more current as the voltage drops.
You should watch the video again, we're not talking about voltage but frequency. It's frequency which keeps the balance of active power in check. Voltage regulation is less critical than frequency.
@@Chopper153 it's possible for a video to inspire other thoughts in people, you know? The delicate nature of our modern electronics contrasting with both the complexity of the grid and how much it has had to evolved since it was first being built.
You really should research grid forming inverter technology. Every battery connection in Australia I work on now uses grid forming technology, which does not synchronise with the grid based on a PLL. Almost all projects in Australia are now also hybrid PV/wind and BESS, meaning many of these concerns are being addressed in all new projects.
I work in Utility Scale Renewable energy, and we also need to ensure grid relaibility for LVRT, HVRT, Harmonics, Frequency etc. This episode quite nicely shared the challenges of large scale renewables.
I hadn't heard of this event before but after watching the video, I'm pretty sure the power plant that experienced the failed arrester was one that I was the Lead Electrical Engineer for during is design and construction back in the early 2000's. I've spent over 25 years designing and building gas fired power plants as well as another 10 or so years developing utility scale PV facilities and this video does a wonderful job explaining a very complicated subject in a way that I think is understandable to most people. Keep up the great work.
I was camping just the other day and my friend who works for GE was telling me about synchronous condensers. Interesting that this video comes out the next day. Was hoping to hear a bit more in depth about how we manage, at least outside of inverters.
From what I recall, in the early days of home solar. There was the issue of these small systems de stabilizing the relatively small Hawaiian grid that all home solar was ordered off the grid until things got figured out.
There was a problem with too many producers on the same local wire pushing the voltage too high, which was fixed by requiring inverters to stop feeding above a certain voltage. I don't think they orderered anything off the grid.
The joy of crating those videos certainly does come across, and seeing someone who's excited to talk about stuff they like makes watching that much better :)
Audio nerd here and you stole the words out of my mouth with that inverter explanation when I thought "That just sounds like the way pulse width modulation works."
Same here! I saw the stair-step waves and thought, "Digital audio!" I never realized inverters are basically 'digitizing' the power to make it into A/C.
Though normally solar inverters are horrid RF noise generators, as most of them really do not have enough filtering, and often enough the connection methods used are not the best from a RF noise point of view. You really need a lot more filtering to reduce harmonics, and such filters at power are both heavy, expensive and tend to run somewhat hot, losing power in the noise by dissipating as heat in the inductors and capacitors. No way to work it with smaller power, though you can reduce emitted noise quite a lot simply by connecting grounds correctly, and using properly RF bonded steel conduits for the cables, and a local RF functional ground as well. But most will not install that, as it adds a lot of cost to the system for little perceived improvement.
Reminded of my first time finding out what PWM is. Had to change the cooling fan motor on Lincoln and I thought it was broke. It wasn't. It was turning at whatever speed it needed to. Once I got it hot it went 100% full blast. Then I found out it was PWM and explained it to the owner, he thought it was broke because it wasn't turning as fast as he thought it should.
Thanks for the excellent video as usual. This is a side of renewables that is seldom talked about outside of engineering circles so I am very happy to see such a good video on the topic. I have a clear memory of my 2nd year power systems class where we were shown a demonstration of how nicely "spinny things" stabilise the frequency when demand increases.
Great stuff! Got a solar install here (in Ireland) which allows for off-grid usage during outages though there is some downtime in the switchover as it has to completely disconnect from the grid to operate. Still a useful option if you can get it.
@@fuzzix Without having any clue about the amount of battery storage of the setup I reckon it could be quite useful. Should be sufficient to run the lights for a while and keep the fridge and freezer at the intended temperature until the sun is back ideally without clouds. I remember a power outage at my parents once during the winter, one complication of that with central heating I had never thought about at that age was that it'd kill that too. Sure the actual heat comes from burning gas but without electricity to pump the warm water around and run some electronics involved with the burner it'll just sit there being pretty. Fortunately it wasn't all that long and they have a wood fireplace in the living room, but older buildings sure can get cold quickly if it's freezing along with a strong wind.
By the way, the reason why they're called 'inverters'. Back in the old days of mercury arc rectifiers, there was only one type of conversion you could do to electrical power: AC to DC. Any unit that did AC to DC was called a 'converter'. Eventually, technology moved on and it became possible to go from DC to AC, the opposite process of what a 'converter' did. Hence, it was called an 'inverter'. Since then, the term 'rectifier' has become common for AC to DC ('PFC' appears occasionally, don't get me started), but 'inverter' for DC to AC has stuck.
I absolutely love how clearly you discuss the challenges with inverter-based resources. I work in nuclear, and have had many discussions about how so many people think we can run the whole grid off wind and solar, but there are so many challenges that spinning turbines help resolve and inverter-based resources would struggle with. But discussions so often devolve into mud slinging about the demerits of each generation type rather than a smart discussion of actual problems and complexities of running a power grid. So I appreciate how you just discussed how it works and is operated in this video!
In Toronto, you can either connect it to a battery, or the grid, but not both. They won't let you use an automatic switch so that you can sell your solar power when the grid is working, but then use your solar power during a blackout.
You get grid tied inverters with battery now, that will charge the battery as priority with solar input, and excess will be fed back via grid tie, and after dark they will run off battery to a point, then leave the battery and run off grid till morning, when they use the solar to charge the battery again. Manage your power use and you really can be a net exporter, never actually drawing power in from the grid.
Utilities increasingly make it difficult to sell power back for any meaningful gain. Solar panels can still make sense, but one really has to run the numbers carefully. Assume worst case scenario for uptime and maintenance. Panels can last decades, but many components won't and are consumables that need to be periodically replaced. To put it another way, figure whatever a solar panel company is promising is fantasy and budget accordingly. As in 10 years payback at most. If it's longer than that, likely best to wait it out for now.
@@ronbennett7885 There are a couple of issues here. First, often solar panels are producing power when there is excess in the system, so it not valuable. But if folks do get high feed-in tariffs or net metering schemes, you essential have poorer people paying for rich folks having solar systems. Third, people scream bloody-murder about having to pay for grid connection fees when they are using little or no power. But someone needs to pay for the grid.
@@ronbennett7885 oh back when I did this, the Ontario government was buying it back from us at 55c/KWh. On a 20 year contract. It's still going! Just can't actually USE the solar power cause the electric company doesn't trust the auto switches and doesn't want me sparking their sparkies.
It seems a bit like the answer in the Edison vs Tesla debate may be shifting. The decision for AC came from a time when phase regulating equipment was only needed in a hundred or so power plants across the country while inverters were expensive mechanical devices, but it allowed us to have "dumb" transformers in hundreds of thousands if not millions of places across the country. However, now this means that we need "more modern" phase regulating equipment with the inverter in each solar, wind and battery installation. And with our goal of expanding renewables, we'll likely see solar installations on the roofs of a sizeable fraction of houses, and thus possibly more inverter based sources than transformers. I wonder if there would ever come a point where it would be cheaper to have built a DC grid instead. We'd need an own inverter on the input side of each transformer, but they just have to be the solid state kind without any synchronization features that make the "more modern" ones more suitable to the current complex landscape, since they can be independent and don't have to "play along nicely" with the rest of the grid. As far as I understand it, the fact that the frequency gets messed up whenever there's a disturbance results from the fact that there's more than one source of AC power on the same wire. With a DC grid, that can entirely be avoided and transformers should work at any input voltage (they could merely output AC power with a lower voltage, while the frequency can be held sufficiently stable to protect appliances). And even if the DC path never gets cheaper, the AC decision was made at a time when only factories and street lighting really relied on the availability of electricity (if they weren't still gas lamps). These days so much more goes down when electricity is out. Not just all indoor lighting, communications, all of our home appliances, and hospitals; but also our heat, water supply, traffic lights, shops, and the power plants themselves. (Yes, there's exceptions whenever outages are too disturbing that backup generators or USPs are kept on standby - which also costs money.) The resilience of not having a single bad law, guideline, or default setting in a series of inverters turn a slight disturbance into a full power outage, might even justify a solution which costs more but isn't so complex that predicting it's behavior becomes near impossible. I mean, even running the self-regulation on the current power grid requires a separate data channel to communicate the current (and possibly even forecast) electricity price. A DC system would in theory regulate itself to some voltage value in the acceptable range, even if you did something as simple as setting a convention how the electricity price varies with grid voltage, possibly even with additional dependence on the time of day to further stabilize the voltage against predictable cycles of demand (and supply).
That moment actually passed a long time ago - all computing runs on DC (transistors are DC devices) so we have to convert AC to DC for every single device which uses a computer, chip or logic circuit, which is everything in (not even modern) electronics. The cumulative costs of that passed the costs of a DC grid ages ago. But there's never been a point, and never will be, when making an entirely new DC grid and switching everything over from AC to DC is cheaper than just continuing to make do with the AC grid we have. Not only do you have to build a DC grid, but you have to rewire every building and replace every device built for AC. Even all the devices that would work better on DC would need to be replaced because they've been built for AC input. The cost would literally be all capital investment across the entire economy for several decades (that's what you're replacing), just to get to the same point we're at today with a DC rather than AC grid. If only Tesla and Edison had a crystal ball that predicted the technical workings of computers - they could have made the right choice, and also invented computers.
I know of a few new factories here in Austria that use two separate power grids. One classical AC, the second one DC to use the electricity from the solar panels on the roof directly.
It's interesting that, DC-sources are causing people to start thinking outside the AC box, and start looking at using the DC directly (pun intended). The simplification it creates has benefits.
I've wondered when DC end to end would start happening (even if in small scale), given that more and more loads either use DC (lighting, electronics), or in principle could just as well use DC or AC, while there's increasing generation that's inherently DC (solar). As an aside, AC is responsible for much loss in the grid (for the same reason that transformers work... induction). More and more long distance transmission lines are DC. So the chain is sometimes DC generation, DC for part of the transmission, and DC loads. I've read that some companies are working on fully DC grid designs. I doubt that it'll become the norm in my lifetime, but I wonder when the benefit of transformers won't be sufficient to justify AC, if DC grid technology gets sufficiently good someday.
@@bearcubdaycare one of the problems is, DC doesn't work well over longer distances, not without some seriously big cables. but if you were going to do certain dedicated things, like say a direct PV connection to an electric hot water heater, then it might not matter as much.
It tickles me to think of fancy new Austrian factories working on basically the same principle as the old camper trailer I used to have (which had an external AC connection for the aircon, appliances, and wall outlets, but ran all the lighting off an internal 12V DC circuit powered by a car battery). ;)
I am not an engineer, but I have a personal interest in such things and have picked up a good amount over the years. That said, some of this was slightly over my head, but you present in such a way that an enthusiastic amateur can gain a greater overall understanding of the situation and the problems that are trying to be solved. *Thank you* for your efforts!
I've also heard about the problem of voltage rise near solar installations- as solar panels doesn't use the concept of power angle like generators, the only way they can inject energy into the grid is by increasing the voltage a little bit. This is fine for a few panels, but in residential areas having large number of installations, each panel has to up their voltage, resulting in significant rise in voltage in lines in the vicinity.
Mostly because the grid was designed as one way, you would need to have a lot more automatic on line tap changers in the residential suburbs to counter this, as they typically are only used in the larger local distribution network for the primary incoming transformers. Local uses mostly off line tap changers, set, locked and left as you cannot change them energised. Automatic tap changers in areas with lots of solar will help, but the expense for the utility is not warranted, seeing as it is not power they are able to bill for easily, and thus low on the priority list to implement. Residential inverters are programmed not to exceed peak line voltage, so the power loss is purely on the solar household side, not the utility side.
Usually there are certain limits to this effect, so that PV inverters will lower power to prevent the voltage from rising too much. In addition, there are actually concepts to counter the effect by drawing reactive power (lowering the cos(phi) of the current injected in the grid) at the same time. I don't know if they are implementet in current PV systems, though.
I dunno about "significant", but your surmise is wrong... grid-tied inverters do use power angle like generators. They push phase, not voltage. But that does have two side-effects which do cause the voltage to rise. (1) The homes are burning less energy from the grid, the lighter loads mean higher voltages. (2) The exported energy has to go somewhere and since it isn't going to frequency that leaves only voltage. Even though the inverters are pushing phase, there will be some voltage rise on the whole circuit as the energy smooths out over distance. But again, "significant" ? Its possible but exceedingly rare, and it will still be in-spec since the grid-tied inverters trip-off if the voltage goes out of spec. A very real problem is that voltage rises in residential voltages feed-backwards through the pole AND substation transformers which multiply it back up to transmission voltage levels. And even when these rises are within specifications, the multiplication on the way back, if the backfeed gets all the way to the substation, can cause the transmission line voltage to go out of spec. This is easily rectified but requires a transformer upgrade at the substation (basically to a modern transformer with two-way sensing and automatic tap control). Fortunately, though, this sort of situation only occurs with a very small number of circuits, so mostly problems develop because utilities are blowing smoke and purposefully letting them get out of control to try to convince politicians to slow solar installations down. -Matt
All that spinning mass is the stored energy to get through transient events. On an airplane I worked on there was a motor generator to both get a particular DC voltage an aftermarket system needed AND it also ensured clean, stable output even if the input AC was interrupted briefly.
An NC (Asheville?) TV station was said to have had a flywheel 'buffer' between the mains grid and the on-site diesel generator. That way, when grid mains were lost, the switchover was not a blackout / brownout event. The spinning flywheel buffed out the transition. Sorry, but I don't remember the call letters. I'd heard this from TV engineers back in the 80s/90s.
I used to work for a Utility level inverter manufacturer and our California customers often opted for the Ride Though board. They allowed the units to keep running for a couple minutes while the grid went through a black out (happened often back then). Fun times!
I wanted to watch this because my favorite presentation at a data conference a few years ago was from an electrical engineer from Dominion Power (my power company). He started with a bit of an overview of electricity and went on to show how his team was monitoring fluctuations in order to enable the power company to better handle the increasing domestic solar installations. But you covered a different angle! Interesting information! also, i loved that talk so much that i asked several other ppl if they liked it as much - they all couldnt follow his intro overview at all - I guess the 2 years I spent as an EE major were good for something.
Afaik in Germany you are required according to vde-ar-n-4105-2018 "Therefore, in the future, newly constructed generating plants must support the grid in case of disruptions."
That would mean more complex and expensive power conversion equipment, but it's what's needed so it can come out of the corner as more than a niche player.
@@SeekingTheLoveThatGodMeans7648 Ehhh, not really: at worst you need a small battery about 20% as expensive as the solar panels themselves, and that's for a much higher level of capability than the regulation mandates. Those mandates are basically all just controller/brain requirements, which may have some effect on what sensing circuits the inverter needs to feel the grid, but nothing that actually handles power. That's the beauty: the normal inverter already needs to have all that power conversion circuity just to feed a reasonable sine into the grid when the sun shines.
I work on photovoltaics and batteries directly, but it's really cool to see the other side, on how people work on installation and application. Thanks for this!
the other reason home solar arrays typically shut down is because they are not allowed to remain connected to the grid in a power outage, and it's easier to shut down the array, than to have able to disconnect from the grid and operate as a standalone system.
@@thedude5040 a standalone system pretty much has to have a sizable battery bank and be able to manage production based on the battery bank. and have an air gap transfer switch to disconnect from the grid. doable, but people haven't started asking for it in enough numbers to inspire manufacturers to go there.
@@thedude5040 it's not going to cost that much MORE than the grid connected system. and a 5 gallon gasoline can is good for about 8 hours, depending on the generator.
It’s not just a challenge to provide ‘fault ride-through’ during transients, but also to provide sufficient fault current at the source, for the generators etc to trip correctly without damaging themselves. If there’s insufficient current available at the generator, as is often the case with renewables (they don’t individually have the inertia of much larger generators), especially when connected over long distances - the protection won’t function correctly and they’ll end up burning out equipment and cables as they unable to ‘disconnect’ themselves. It’s why many grid connections of renewables require large synchronous condensers or similar, to provide the immediate energy/inertia required for protection at the generator source to operate quickly and correctly.
I didn't understand why grid frequency drops when there is more demand than load, but this makes sense: >whenever there is increase in active power requirement, kinetic energy of rotor is used to supply this increased power demand which lowers the speed of rotation and results in frequency drop.
THIS‼️ I am really confused as to why Grady did not discuss where the added power comes from. My assumption is the same as yours, but I did not hear it in the video. And, I am shocked that I had to scroll through 100s of comments to find a mention of this. I feel like I must've missed something.
Practical Engineering prepared a good video on the subject. The take aways are that with the present technologies the grid will need conventional rotating machines to maintain the frequency and the voltage support for the frequency following inverters to function. Following the technical literature there seems to be a limit for renewable generation with frequency following invertors at 70% penetration due to dynamic characteristics of the grid following a major disturbance like a fault. As cited in the video there is a need for inertia and the ability for reactive support from generating sources to ride through the disturbance. One must remember that the invertor-based devices can provide some of these requirements, but there are limits within the controls for them to be effective. For the grid operators the information of those limits needs to be well understood. There is plenty of work to be done for achieving the goals.
In Finland, it took over a decade to make necessary modifications to the grid, and we are finally in expansion upgrades only mode. And we had a state of art power grid already in the 90s.
commendable civil eng. channel, deep dived into electrical/electronics, likely required consultants for this video. however, it was only difficult for a brief period when renewable energy is rising and battery energy storage isn't common affordable yet. these days, this problem is virtually eliminated by BESS and smart grid systems.
Just started to explore Grid-tired solar system. This video came a a good time. I was discussing with our Solar vendor about having the inverter continue powering thru a grid outages... and the Solis smart inverter could not do it. Yup, knew of MPPT, and how Solar Generation is not as smooth as most people think that it is, passing clouds could drop the generation just like that, and it will cycle thru generation at Peak and trough...not an easy task, trying to do power generation, and still have to follow the grid's frequency and voltages.
With a grid tied inverter, increasing or decreasing the voltage will just push or pull VARs from the network. It is actually shifting the phase of the inverter ahead of the network that will push power out.
It's interesting to think about how solar and distributed inverters could be used to provide half-cycle ride through and power factor adjustments. Seems like just changing our thinking we could use these to our advantage.
@@huckleberryfinn6578 That's Adam Smith air-quote "capitalism". It relies on this underlying assumption that the best solution will always be the most profitable. It doesn't take a Nobel laureate to figure out that assumption just is not the case. PG&E is a classic example. Much of their infrastructure is decades past its rated service life. But upgrading and replacing it cuts into profit, so instead they cut power to their customers and burn down the countryside whenever there's a light breeze. If you care to educate yourself on some of the thicker weeds, you should look up concepts such as price elasticity of demand and hydraulic despotism. These two principles are the Achilles heel of capitalist ideology.
This was mentioned, just briefly, in the video - some inverters can provide what I think he called 'virtual inertia'. Either boosting the front of the sine wave or loading it down if the frequency drops or peaks.
I remember that balloon escape from East Germany back in 1979. That was some pretty creative engineering just to get it to fly in the first place. But then to actually make it across the boarder even with multiple mishaps along the way, is amazing. As a student in 1972, I got to spend a summer behind the "iron curtain". Those governments didn't fool around, there were police everywhere.
Another example of grid frequency blackout is the eastern seaboard blackout of 2003. The grid was running at capacity when a powerplant tripped, and 8 states and southern Ontario tripped offline. Power lines have their own internal resistance (called volt/amp reactance or VARs), and resistance converts current to heat. The more current you try to push down a line, the more heat created and the line starts to expand and sag. You can actually see this if you photograph a big power line from the same spot during high and low power demand periods. In the case of this blackout, a line heated up and sagged into a group of trees that hadn’t been trimmed. This caused a ground fault, and a protective relay at the plant did what it was supposed to do and opened the generator breaker, taking the plant off line. This caused a drop in frequency at nearby powerplants forcing their under frequency protective relays to open their respective generator breakers. This lead to more and more areas where grid frequency collapsed, causing a cascade failure that took out most of the northeast grid before it could be contained. The next problem was how to bring the grid back up, and it was caused by people. They wanted to know when power came back, so they went around turning on lights and radios to alert them when power came back up, creating a huge potential load. Anytime a powerplant tried to close its breaker, it would just get dragged back off line. Since the grid was dumb (no remotely operated switches) in 2003, (most of it still is dumb), utility workers had to go out and manually open switches to “island out” areas small enough that the generator governors could react fast enough to the sudden load without tripping. At that point, the people with power would go around shutting off all the lights and radios, lowering the load in that island. Then the next island could be connected in and so on. It took 7 hours to reestablish power to most customers. It was several days for some customers. At no point did any equipment fail. It all did exactly what it was designed to do. The point of failure was the lack of maintenance in trimming trees (exacerbated by environmentalists trying to prevent tree trimming). The grid operators in Ohio were criticized for a lack of understanding of their section of the grid and improper response, but in fact, understanding grid operations requires being able to do the calculus in your head. A bunch of business majors had no chance.
That was really informative. Thanks. I’ve played around with a couple of 400 w solar panels and have run across a lot of terminology that I did not really understand till now. Funny how a small scale solar setup can help make these huge installations easier to understand
Ah, I was wondering why we had a power cut... when we had a power cut. Figured there was something about our solar configuration that meant it couldn't work as the sole power source for the house.
Grady did not cover another reason solar systems turn off when there is no utility. If your solar system is feeding power into a utility that the power company has turned off for repairs then you are energising cable plant that is expected to be de-energised. This can be dangous to linemen. Remember transformers work both directions.
There are three kinds of solar systems. Off grid, make power all the time, can't be connected to the grid cause it doesn't sync. Grid tie, which does sync to the grid, but when the grid shuts off, it MUST shut off, both because lineman don't want to get zapped, and because you will never own equipment powerful enough to carry the whole neighborhood. The third is a hybrid, which can do both, because it automatically disconnects from the grid, and then turns on(often with multiple pieces of equipment). Tesla has a system. They used to have a Tesla gateway, which was the switch that disconnected you from the grid, and let the powerwall know it can turn on. They might have built that functionality into the Powerwall 3, not sure, but their new setup isn't approved everywhere yet. Enphase has a system. SolArk sells a hybrid inverter. Point is, if you want your system to do it all, you have to design it that way, and pay for the parts. I just have a grid tie from 2017, but I want to eventually be able to do it all.
Solar is excellent for future-built homes. You add DC outlets (or USB ones) designed specifically for electronic devices, power tool battery chargers, or electric cars so you can directly power them without having to invert it to AC and back again.
I think this would work well. Look at RV's, the have 12 volt, 120 volt, and gas systems, why not our home and work places. lights and electronics could all be DC.
I like the concept in that most electronic devices, if you look at the little wall adapters for them, run off of 5VDC or 12VDC. Its a whole lot more efficient to buck-down a modest DC voltage then to convert to AC and then back to DC. But there are some warts. The biggest wart is a safety concern. You can't use AC breakers with DC voltages and you also have a problem with ARCing when wires break that you don't have with AC (because with AC the current crosses zero 120 times a second). So even though we can have in-house DC wiring, its a bit more dangerous when faults develop as the wiring ages. As such, I think DC circuits are an interesting concept, but amperages would have to be severely limited and electronically controlled., possibly with an insulation resistance check as part of the safety device requirement. I think something like .... well, for house wiring you don't really want 12V. You want something like POE+ (power over ethernet) which is an actively negotiated protocol which also actively checks available wire pairs continuously. It runs at roughly 56VDC. Almost impossible for POE+ to cause any sort of fire. It is typically limited to roughly 60W over four pair which is barely 15W per wire pair (0.35A or so per wire pair). There is also POE++ which supports higher wattages by using higher voltages. -Matt
I'm still waiting on reasonably priced DC breakers to actually use that (not at my home, because I don't have a good place to put solar), particularly actually for a mini-split AC/heat pump.
@@junkerzn7312 Modern AC breakers actually have substantial DC capacity (and rating!), it's just that you can't really handle more than 60V with a single contact MCB. There are low voltage DC standards in place currently for e.g. iirc a bipolar +-125V system without locking connectors, and an iirc about 380V system with locking connectors that's been deployed in data centers for a while now (their DC is their battery backup bank; they skip the inverter of their UPS).
@@namibjDerEchte Well, I've got like four different brands of DC MCBs in front of me and the unpolarized ones are typically rated to around 110VDC and the polarized ones up to roughly 250VDC, I think I have one PUFA here somewhere with a 1000VDC rating. Generally speaking, the DC MCBs with higher ratings tend to be polarized. The DC MCBs with lower ratings tend not to be. But this is an area that get DIYers into trouble all the time. Either they buy polarzied breakers and don't hook them up properly, or they buy polarized breakers and don't realize that they can't be used in a battery circuit where current can flow both ways (charging and discharging). OR they buy a non-name breaker and just assume it is polarized or unpolarized from the markings, not realizing that most polarity markings are completely meaningless as a "tell" on the type of breaker. Most AC breakers do not have DC ratings and definitely cannot be used in DC systems. Not sure where that came from. Expensive high-end breakers will often have both AC and DC ratings but most run-of-the-mill breakers that people buy... those almost never do. The construction is too different. Typical AC breakers use relatively large GAPs to break ARCs for example. And also often don't have arc extinguishers. But that method relies on the current crossing 0A 120 times a second (AC waveform). DC with a breaker like that will build a lot more carbon up on every trip until the breaker fails and fails to break the ARC, then catch on fire, Typical DC MCBs use narrow gaps and walk the current up to the ARC extinguisher. That can be done both polarized and unpolarized. Polarized breakers can walk the current more quickly and more reliably and tend to have higher voltage ratings. Unpolarized breakers still walk the current away from the contacts but don't have magnets near the contact point forcing it to walk faster. But being unpolarized, they can break current going in either direction. -Matt
Actually most wind turbines usa an asynchronous induction generator. They drive this with a fixed gear ratio to maintain rpm just above syncronous speed. Thats why they always turn at the same rate, no matter how strong the wind is. When the wind is too slow to generate any power at this speed they turn all the way off.
Let me just say that I am a big fan of your videos as I am an electrical engineer and I love your deep dives about practical engineering. I think overall, you summarized the topic of challenges regarding inverter-based resources quite well because lack of inertia, ride-through requirements, and protection issues are among the the top 3 challenges for sure. I am only commenting here because this is the only topic you covered that I have some expertise with. I really do not want to nitpick here but I felt like there were some small errors with the way you explained how grid-tied inverters work because once synchronized with the grid, they do not vary their voltage to control how much power is flowing into the grid because the grid will dictate a constant voltage and frequency. They vary the phase angle between the inverter voltage and the grid voltage to control the current flow and thus the power flow to the grid. The voltage amplitude stays constant and that is why the graphic is also misleading showing the voltage waveform going up and down while this is inaccurate. The explanation of how inverters work was a bit oversimplified for my taste of course and you showed a graphic of multi-level inverters instead of pulse width modulation which looks a bit different but I think that is okay since you are explaining the concept to a general audience. Again, I was so excited that you finally made a video about a topic that I care about and know a bit more about haha. I am not pointing this out to criticize your efforts but because I feel like it is important to try to give accurate information even though it is challenging when you try to simplify it to be consumable to a general audience. Keep up the good work :)
@@Shaker626 , grid level semiconductors are not using shrinking to improve economies of scale. They're much larger due to the power they're handling, so Moore's law has little impact either way.
@@JimBob1937 The wafer production backend is a big chunk of the cost of those chips (their fabrication is not as expensive, however). Moore's law was the main driver for making wafers cheaper, which is not very likely to proceed at the same rate. It's unlikely that power transistors will get much cheaper than they are now with the IGBT.
@@Shaker626 , wafer production is based on growing a cylinder from a seed crystal and then slicing and polishing it up into a set of wafers. This has no connection to Moore's Law. Moore's Law is more about shrinking down the transistors via smaller and smaller interconnect technology and better and better photolithography (most recently EUV, due to the shorter frequency giving better detail resolve). This in turn ither allowed for an increase in transistors per fixed area of the wafer, or for less wafer area for a fixed number of transistors. This allowed the wafers to be better utilized. However, grid level semiconductors are not concerned with size as much as power handling. Nothing of Moore's law directly affects them, since it isn't an area they're concerned with.
4:37 .. These are still popular today .. There are several, single to three, "rotary phase converter" companies, and the old "Redline" 12 volt DC to 20 volt AC "rotary converter" is still used today.
Grady, I'm sorry to do this to you, but it kills me to see smart people doing this. At 2:18, it's either "comprising" or "composed of." "Comprised of" is not a proper usage of the word. Just a PSA! Big fan though and I love this topic. These videos where you go over events are my absolute favorite. You're a clear, concise storyteller and super great.
This was a great episode. The comments were as informative as the content. The person from the EU who have enough solar in neighborhoods that when the grid trip off, the neighborhood may not due to backfeeding was something I never even thought about. My only real comment about it was There Were No Googly Eyes On The Demo. There are always eyes on the parts of the demo. I would have thought the ideal place was on the hard hat sitting on top of the scope, so it could watch that perfect sign wave from the inverter. Thanks for another great episode.
It's the American way, maaan. Like you wanna go all socialist on us? Ewww.....and so on and so on....sigh. Thank you Gordon Gekko, you've f***ed the country good and proper..
I know right? Why TF are all the power companies saying they need public money to upgrade the grid for EVs when they stand to sell two or three times as much power in the end?
I'm jealous of your ability to vulgarize these concepts to a larger audience, grid forming is so hyped right now ! These backyard demonstrations are also very cool :)
1. Attaching any (heavy) power equipment to the grid (i.e. an interconnection) requires permit from power utilities. 2. A badly planned interconnection will cause equipment and possibly grid failure, ending up "in smoke".
@@Speak4Yourself2 Thank you. I still don't understand why he calls it "making an interconnection" though, like how can he create one by plugging in a device?
In his demonstration, he showed the output voltage from his inverter and the voltage from the grid. You can see that the voltages on the screen are very unsynchronized. Practically, when power plants connect their generators to the grid, they have to adjust generators' voltage and frequency to be synchronized with the grid. On the screen, the frequencies of both signals aren't equal as you can see that the signal moves relatively to the other signal. And voltages are even worse. One is sine wave while the other is step. If you connect them together the difference of the voltages will cause high amount of current and thus cause the smoke.
@@sebastianelytron8450 Yes, connect the inverter's output to the grid. As you can see on the screen, the inverter tries to make its output waveform to be kinda step waveform while the grid is sine waveform. Two outputs with different voltages should never be connected because those devices want to regulate their outputs to be what it's designed for. (This applys to everything, USB, HDMI, etc. You have to connect from one device's output port to another device's input port.) Two outputs trying to make difference voltages will cause high current as the interconnected wire resistance are very low and I=V/R. It's basically a short circuit. I think you should check out the video titled "What Is A Black Start Of The Power Grid?" on the channel starting at around minute 10:00 to understand the term "synchronization" and come back later if you still have question.
Outstanding description and visualizations for those of us who are not engineers. I'm a laymen selling radio communications services into utilities and this kind of information is soooo good!! Keep it up.
Great video, however there's a slight error at 6:14. You mention pulse width modulation with incorrect context: PWM has to do with the ratio of "on time" to "off time", most commonly used for pure square waves, and doesn't have anything to do with the number of steps used to approximate a sine wave. The term you're looking for is "switching frequency", analogous to the number of bits in a DAC.
I built a solar powered kayak by putting a large flexible solar panel on a PVC adjustable frame. It acts as a sunshade, and powers a standard 12 volt trolling motor via PWM to provide the optimum panel loading. The fun part is this is done manually to find MPPT, so changing course or a cloud coming over requires PWM adjustment. Like sailing, where you need to constantly adjust for the wind.
The power conversion..... invokes memories of selecting Uninterruptable Power Supply units for some computers many years ago. Cheaper models put out a "stepped sine wave" and better units put out real sine waves. One section in the video explained details that I never knew (all I knew for UPS selection is that real sine waves were better). Thanks!
Some of it is just for safety, but 90% of the problem with solar in my state is political - the utilities have complete control of the legislature and the laws are punishing for anyone who isn’t them
This gives you an opportunity. You can get retired panels and connect them to an off grid load. In Phoenix I would put retired solar on the roof and connect it to a mini split for free air-conditioning
@@TimHaywardRetired panels? Never heard of it, but now I'm definitely curious! Why do they Retire them? Do they need some type of maintenance? Where do I find these?
Thanks for sharing your ideas, thoughts and videos. As a retired senior system operator I always find your videos interesting and usually spot on. About the only difference is your use of scientific units. In the bulk power industry we usually talk in terms of megawatts for communication clarity. 500KW would be 0.5MW. Or my import limit would be 1200MW not 1.2GW. I have always enjoyed when people build simulations of larger systems, really liked your demo. Was reminiscent of a simulation I once saw using WW2 era aircraft motor generators. By removing the shielding you could see and manipulate the rotors. By tying them together you could simulate how the emf force thru the wires tied them together. Since they were permanent magnet motors. Hand turning the rotor on one generator would cause the other generator only connected by wires to turn as if by magic. Another technology you may find interesting to look into are VFT aka variable frequency transformers. They are made by modifying a three phase motor with the stator windings as one half of the transformer and the rotor windings as the other. By controlling the motor input shaft with a stepper motor you can vary the phase between the two systems and push power one way or the other dynamically. An example of this can be found at the cedars station between the NY and HQ power systems. While learning of the technology when they were training us for its introduction to the power grid a fellow operator recognized he had seen a similar setup earlier in his career at a smelting plant. They used a hydraulic cylinder to push and pull on a crank arm attached to a motor input and rectifiers on the output. Making an industrial sized variable DC power supply.
📖Signed copies of my book, Engineering in Plain Sight, are back on sale! store.practical.engineering/
🌌Get Nebula using my link for 40% off an annual subscription: go.nebula.tv/Practical-Engineering
Oh snap ... I've never thought of distributed generation solar increasing rocof. It's a lose lose situation for grid tied solar. The next Gen nat gas turbines (GE and Siemens) seem really cool, and a good mitigation for this issue
Hey! Solar interconnections are what I do for a living! Always great videos on the power grid. You reflect the info very accurately!
I wish you had a media mail shipping option for your book. I’m only a few hours away in DFW, and the cheapest shipping option right now is over $12.
O-scopes everywhere are jealous of the hard hat you provide to protect yours.
❤
In early 2025, the Baltic states in North-Eastern Europe are set to disconnect their grid from the Russian power grid, and connect up with European power grid instead. This might sound like it's simply some sort of resynchronization task, however the reality is much more complicated, requiring the Baltic states to build out a fair bit of extra infrastructure and take on the task of maintaining grid frequency where before, Russia would have been doing so. This might make for an interesting video.
They better hurry up
I live at a junction between two power companies' service areas, and they have the ability, in an emergency, to connect one to the other, but there is the catch that there has to be a transformer, there, because the operating voltages don't match - and there's no synchronization, and there's a meter to measure the power that crosses junction.
Should've been done years ago tbh
@@kenbrown2808 Brazil has the same issue, along with Japan, and built HVDC systems that are both frequency changing, and bidirectional, so that power can flow either way, along with changing the frequency of that power to suit the local grids as well.
@@personzorz why? 😂
I work as a lineman in northern EU, and an interesting effect we've noticed with solar is that sometimes the inverters won't disconnect from the grid during a power outage. We've ran into this when doing service work on pad-mount transformers, when we open the switch to de-energize the transformer supplying a neighborhood (in our case an average of 150 households/transformer) with a lot of solar, on rare occasions it will remain energized, backfed from the households solar production. This happens because in the right conditions the solar panels will produce the same amount of power as the households consume, so there's no current flow through the transformer. So, when we open the switch - as far as the inverters are concerned nothing changes, so they remain on and backfeed the grid. It usually doesn't last very long since when the balance shifts they'll disconnect but you could still get a nasty surprise if you're not careful.
Ah now this is technically illegal, all devices that back feed to the grid must have anti-islanding devices fitted. Well in France and the UK. When I worked for RTe (french transmission operator) there was talk of taking some people to court etc about this issue. The same goes for houses fitted with a gen, it must separate from the grid when the grid goes down.
Woah, that's an interesting fault. I've never heard / thought about this exact configuration, but yeah, they would stay on as long as that configuration remains stable...
@@matthewmaxwell-burton4549 Yeah we have the same here but the issue is them not being able to tell if the grid went down, as long as the flow stays balanced at least
>lineman
Lying American. We do not have lineman, as we do not install our lines above ground like savages.
I learned today that new installation of a grid-interactive inverter must: have ability to throttle the power and also be able to shutdown the Inverter power and control the scheduling of when the power is used/delivered and the quantity on command from the utility. I imagine the direct command & control will tend to resolve this accidental self-supporting meta-stable microgrid situation.
This was interesting. I was a nuclear trained submarine officer and served on an improved LA class sub for 3 years. Going through the nuclear training pipeline, you learn a lot of this stuff. In fact, the crude demonstration in the video of a drill turning an AC generator is something we had two: the "ships service motor generator" (SSMG -- everything needs an abbreviation), which were approximately the size of a midsize vehicle. They converted power back and forth from the AC to DC bus, where the DC bus was powered by **VERY LARGE** lead-acid batteries in the bottom of the sub. Very interesting stuff. The control room ("Maneuvering") had the panel for managing the electrical buses, and the operators manually would synchronize various bus frequencies ("slow, in the fast direction" for the incoming bus, and throw the switch for the breaker at approximately the 10/11 o'clock position to shut the breaker at 12). Depended on the operator -- not automatic. I believe newer subs use a solid state inverter system and not the SSMGs now. Anyone really interested in hands on, great education in this sort of stuff, either as enlisted or officer, join the Navy and go to Nuclear Power School and get to a ship. I wasn't an engineer in college, but you more or less become one working the engine room of a nuclear submarine.
That's really cool, thanks for sharing!
i get seasick :/
I taught NNPS when it was in Orlando. After that, I designed and built the electric plant trainers based on the S5W (Bluefish, et. al.) plant used in NFAS back in the late 80's. Modeling the SSMG was one of the more challenging things I've ever done as an engineer. I used capacitor charging curves to emulate the starting surges that occur when the resistor sequence is pulled as the rotor speed comes up. The ships battery was a 3 D-Cell actual battery pack, that was literally charged and discharged from the op amps of the model. Overall that project was the best engineering task I ever undertook, and got a Westinghouse Signature award for the work. But they wouldn't promote me ahead of "years in grade" ( just like the military!) so I quit. And BTW both Bettis and KAPL were absolutely bursting with "engineers" that couldn't design a resistor divider.....but they wrote beautiful reports to be sent along to Navsea-08 to impress them with how much was being accomplished. 😮
@@jeffferguson4632 Oh, that's neat! I was one of the last classes through Orlando NNPS in 1997. Your comment about the resistors in the SSMG startup is memorable. The DC startup on those machines caused the biggest transients on the electrical panel -- always something to watch when that happened. Plus, I recall a teachable moment about good engineering practice when we restarted an SSMG (following routine maintenance) while in an unusual electrical lineup. That's quite a thing to have been involved in the engineering of. They worked well from my perspective.
I was involved in the early design of the Moored Training Ship setup in Charleston....when the simulator was located in the trailer. You wouldn't believe the arguments we had about putting a "head" in the trailer. If you went to MTS Charleston and bitched everytime you had to go back to the building to take a leak....well...I tried.
I am a test engineer for high power frequency converters and be build our "VSD"'s not only for grid injectors like wind turbines (exactly like in the video), but also for applications where you would not suspect them. Is is for example sometimes easier to let a water turbine run at a non-constant speed and run all that power through a VSD, which constantly synchronizes to the grid, than try to maintain a constant speed with the turbine. The are also used as a soft starter for these very big motors and sometimes, they can even reverse the flow of energy to pump water back up into a lake for example. We are talking from 500kW to 100MW here.
Great video!
I worked at a factory that ran 24/7 and consumed megawatts of power. We were looking for backup power solutions. The problem was the few seconds backup generators take to start. There were "batteries" that consisted of carbon fiber flywheels in a vacuum spinning at 500,000 RPM. They could deliver 100% output (multiple megawatts) within 10 cycles, and for up to 10 seconds - enough time to start a generator kept in warm stand-by. Perhaps technology like this is what we need to stabilize the grid.
The UK is spending billions in building flywheels to pick up the janky electricity from solar and wind and add a modicum of control to it.
Or we could just use nuclear 🤷♂️
There are a lot of companies looking into grid scale flywheel energy storage. It's a valuable option and one I could stand to learn more about. But I think the biggest problem with any **grid scale** energy storage solution is cost. I'd love to see a practical engineering deep dive into grid scale storage and fly wheels.
I understand that flywheels are already used by various utilities for this sort of purpose, providing inertia at a moment's notice. Wikipedia at least says that there are grid-connected flywheels in New York, Germany, and Ontario. Would be nice to see more about how these sort of utilities work!
Yeah, I am hardly an expert, I just remembered the Capex research I did for that project a decade ago. The thing with flywheels is stored energy goes up linearly with mass, but squared with rotational speed. So a 1 pound carbon fiber disk spinning at half a million RPM can potentially store more energy than a multi ton wheel spinning standard generator speeds like 1800 RPM.
Chances are people in the industry know much more about the details than I do, but occasionally good ideas slip through the cracks.
Flywheel storage solutions are interesting. I work on mission critical power systems, and most of the time a battery based UPS is used. For certain loads, especially at a factory where there are huge motors flywheel based solutions can be better
I'm an engineer for a utility-scale solar developer and this is a fantastic summary of the challenges we face and design for.
My issue with solar: all we are doing is tearing down forests to build them… lets destroy more habitat so we can build solar power, yay!… looking at you australia
Tree clearing is expensive, slow, and complex. It’s a last resort and doesn’t represent the typical project (at least here in the US).
Just out of curiosity do batteries help "clean up" the power from solar? I've heard one issue is the power they put out can sometimes be "dirty" and cause issues. I'm thinking of solar at home but I've heard those systems can sometimes cause issues for peoples neighbors. I'm not sure with the inverters that people put in along with a battery backup can sorta smooth everything out so the power being output is clean. From this video it seems like the inverter might actually be the issue. As a side note would flywheel storage being added to larger renewable systems help with this or is that still needing to go through the inverter? The flywheel system might not be actually used for storage but it could just smooth out the output? Seems like wind might not have the same issue though since the turbine itself almost acts like a flywheel.
@@pin65371 Batteries basically use the same inverters as PV. Either literally the same inverters (if DC coupled) or models that were likely derived from PV inverters (if AC coupled). Either way, I disagree that the output from PV facilities is "dirty". As the video mentions, it can output a perfect 60hz using electronics alone. They can also be called upon to inject or absorb VARs to help strengthen the grid. The real issue is that renewables are intermittent which can make load balancing challenging for grid operators. Batteries definitely help with this, as they can help firm or smooth the output of the plant to avoid fluctuations. I've never heard of flywheels being used alongside PV, but some companies are looking at alternatives to batteries, such as hydrogen electrolyzers to "store" energy as hydrogen.
@goldenhate6649 That is far from what we commonly do in the US at least for my EPC. Additionally when there are protected areas or wildlife we often have to have a dedicated biologist on site. SWPP is also huge to ensure runoff does not cause problems for thr surrounding area.
the final project for one of my classes was to design a solar installation for a factory, we had to calculate power usage from power bills, design a capacitor bank for power factor correction, find out how many panels, in what arrangement, and at what angle and distance from each other they had to be installed, pick an MPPT and inverter, and design a battery bank.
i couldn't finish in time, it was a nightmare, but at least it taught me that people way underestimate how much work goes into installing solar
Anyone who has tried to design their own solar panels and inverter setup for their home knows that its a complex process that has you reading datasheets for the panels and the manuals for the inverter!
@@tlangdon12I ordered a kit online in 2017, 20 Astro Energy 310W modules, DC optimizers, and a Solaredge SE6000A-US-U inverter, Iron Ridge XR100 racking, and connected it all. I got a Bosh D-Tect 150 to help me hit the rafters through the roof with the 5 inch stainless lag screws holding the flash feet. Making 4.47kw right now. It's all tested by UL. There was a little paperwork to get permits, and pass inspection. Your backfeed breaker has to go in the bottom because the solar adds to the amount of power you can pull out of the pannel, and you don't want it to add that to grid power and exceed the capacity of the bus bar. If you put it in a subpannel, same thing, and it needs to be in the bottom of the main pannel. It's not a beginner project, but totally doable if you are motivated.
Complexity varies. I just pluged microinverter to grid and screwd panels to frame and done. Nothing complex on simple house system. Not a single thought about power factor correction. Its how complex do you want or need to make it.
Wait, what? Was this a school or uni project? Because this is definitely too much work for the former...
@@TheMightyZwom Almost certainly a university project, you would be happy to learn about solar power at all at some schools.
I learned all about MPPT while shopping for solar charge controllers. No tutorial I found actually showed the same panel in different conditions, they just said that the MPP changes. Your test results were nice to see, and helped with understanding. Thanks!
The Odessa event was (edit: contributed to) by a particular phenomenon called Cessation, which is not an unexpected behavior of nebulous algorithms. Designers put it in intentionally to protect the hardware without violating the grid rules at the time, which say when a disturbance happens, you must keep your output breaker closed to help the system (low voltage ride-through.) But there were no requirements that your unit continue to output (real) power. Solar units therefore would drop their power output to nearly 0 to protect the hardware from harmonics generated by the panels trying to push power into a disturbed grid. There was also an automatic timeout on this cessation where the panel would not ramp up until the disturbance had passed for some time. This was often baked into the hardware and many farm owners didn't even know the behavior existed.
6:30 this explanation would have benefitted a lot from showing a raw PWM signal, then showing a capacitor low-pass filter wipe across it to smooth it out
This reads like a case of a disconnect between the objectives of the people writing the grid connection rules and the people implementing them?
@@rowanjones3476 eh, it was more a case of hardware designers not being aware of WHY particular requirements exist. They met the letter of the law while violating the spirit of it.
@@marshallc6215 As a hardware designer, this kind of thing drives me nuts. We're always having to go back to product managers and ask WHY they want things on our boards. But I understand not everyone feels like they have that ability, or perhaps are jaded after being told "just build it to the spec and quit bugging me!"
The key thing here is that the inverter has to push real power. If all you needed was a "sine wave" then you could produce one with a DAC and a simple capacitor-based filter. The problem is that a DAC is a resistive device and can't actually push any real power (it would melt into a pile of sludge in milliseconds if it tried), so the sine wave is useless for inverter purposes.
For the same reason, using e.g. a transistor type of circuit to amplify such a loe-poert sine wave won't work because the transistor is a diode-like device and the voltage difference is dissipated as heat. The transistor can't actually amplify a DC input and produce a ton of power on the output that looks like a sine wave either. It would instantly melt.
Same thing for a PWM signal. The PWM signal itself can drive FET gates, and the FETs are what switch the real power from the DC or rectified source. But a capacitor on that output won't produce a sine wave, it will just melt (more or less instantly) because it is essentially short-circuiting the FETs. Again its an issue of the power flow. A capacitor that isn't doing any work won't melt, but in this case the capacitor is doing a massive amount of work and the FETs are basically pushing incredible currents into and out of it because it kinda looks like a short to the FETs (hundreds of amps.... 1000VA for each real 1W of power).
So the only way to really produce a real sine wave with the desired amount of power is with magnetic components... inductors and transformers. The PWM can drive the FETs, the FETs run the real power through an inductor to convert the medium-frequency PWM signal into a voltage, and that runs through a transformer to produce the desired output AC voltage. The inductor and the transformer both filter what is essentially a huge HF mess of square and sawtooth waves into something close to a pure sine wave and a single capacitor on the output gets rid of remaining HF artifacts.
-Matt
@@junkerzn7312 Ehhh, a modern cheap grid-tie inverter for areas that are 3-phase would just be a triplet of half bridges made of SiC FETs (rated to 3x the phase-to-neutral RMS, measured across the entire half-bridge), with an LCL low-pass between the switching node and the grid-side phase wire, with the PV strings MPPT boosted up to said DC level (e.g., in central Europe with about 220~250 V, we'd be at about 750 V for the DC bus). At larger powers you will want more of the SiC switches in the 3-phase converter section, because the extra cost to control them all separately will be worth the savings in the LCL low-pass filter that theoretically losslessly eats up the PWM's higher-frequency parts only letting the low-frequency smoothed sine through. But you need around 50-100 kW (well, I guess, kVA) before that should be worth the complexity.
power electronics engineer here: Good video. I do have a few additional thoughts, though:
1. You said a home PV system cannot provide power during a blackout. At the end of the video you talk about grid forming inverters being able to support e.g. a backstart or island grids. The good news is: Those systems do exist for home PV applications. They are usually more expensive, but can supply power during a blackout. (There are different types being able to do different things exactly - I'm not going into details.)
2. One problem not addressed in the video (or maybe I missed it?) is the fact that mechanical generators can provide high short circuit currents that will trip breakers (not the ones in your house, the large ones being part of the power grid). Since fault currents have always been high in the past (something like 10x the maximum normal current), circuit breakers were designed to not only handle them, but also require them for fault detection. But: Inverters don't deliver such high surge currents, even if they stay active during the fault event. They generally limit their own output current. While a mechanical generator can survive a current surge into the grid, a - for example - PV inverter couldn't (hence the current limiting). This is different from fault ride through (FRT) and is an additional problem. IMHO the solution to this is more modern protective equipment that will detect faults differently and trip not only during high current surges.
3. What you're showing at 6:06 is NOT pulse width modulation (PWM)! This is (kind of) what a multi-level inverter's output voltage could look like. A PWM still only has two voltages: plus and minus dc voltage (in the case of a two level inverter, that is), but switches between them rapidly, so that the resulting voltage resembles a sine-wave after filtering.
Grid operators usually don't want solar supplying power during a blackout. At best it complicates the process of reconnecting power, and at worst it makes cutting power for servicing very difficult.
The problem with demonstrating PWM is that the fundamental frequency of the modulator is usually off into the hundreds of kilohertz, but the frequency of the final output is in the tens of hertz. This makes is it impossible to show both parts at the same time, The levels of zoom are just too far apart. I think showing that off would have made the video too complicated for its intended audience. Also, he didn't say that what he was showing was PWM anyway. He said that it was multilevel output, although I'll grant you that describing PWM as being just like multilevel but with more switching doesn't really make any sense. And, the way that he describes it sounds a lot more like pulse density modulation, which is quite rare, as far as I can tell, but, in my opinion is easier to explain than PWM. PWM is just way more common, probably because it's a bit more simple in how the algorithm that drives it works.
I mean, if you know all this stuff you’re watching this mostly for entertainment purposes anyway. If you don’t know this stuff, that’s outside of the scope of this video. I like these videos because they give a broad overview of a subject.
1. He said "most" residential solar systems can't provide power during a blackout, which is true. You need a Tesla Powerwall, a Generac battery/inverter system, etc. to have both the storage and the grid-forming inverter to do that, and you need an automatic transfer switch to prevent you from energizing the grid during a blackout.
3. He specifically said this was from a "cheap" inverter (i.e. "modified" sine wave), not from a PWM inverter. There are multi-level PWM circuits. In fact, all 2-level 3-phase PWM inverters (which is most of those described in the video) are effectively 3-level because of the 3-phases. A 3-level 3-phase inverter is effectively 5-level.
@@vylbird8014 A system providing your house with backup power during a blackout does not provide power to the grid! That would be a different thing.
I’m so happy the idea that “magic smoke” is what makes electronics work, is so ubiquitous.
Magic smoke isn't what makes it work, it's what happens when it doesn't work. Something underneath blows up, "guess it was magic."
@@gay4femboysElectronics work because magic smoke is added to some rocks in the factory. When the magic smoke gets out, the electronics stop working.
@@gay4femboysNah, if you actually know how electronics work, especially modern small ones, you'd be blown away how much of a miracle that things work at all.
Beyond walking in a knifes edge, you're walking on an edge literally hundred if atoms thin.
So many insignificant things can make it all go haywire, our maths are amazing
I'm a retired EE, and the sight of magic smoke got me into it decades ago. Still release some, in my hobby time, now.
Don't let the smoke out!
I worked for a utility company a while back and I have to say this is an incredibly well written video
What are your thoughts of using natural gas plants as an in-between for power plant replacement in the near future?
@@Mcfunface that's ok for filling in the gaps right now while long duration storage is developed. But we have a lot of simple cycle capacity throughout the US so I'm skeptical of the need to build out a whole bunch more.
Really what's needed more than new gas plants is better transmission between ISOs/RTOs and a simplified process for getting lines approved and built
Another old, now resolved problem with PV inverters was that in case of an over or underfrequency (I'm going with overfrequency for now), not just one or a couple inverters starting dropping their power but every single inverter connected to the grid. In the early days the power reductions level were hard set which led to situations where an overfrequency triggered thousand of inverters at once to drop their output, causing a sudden drop in the frequency, which then again led all the inverters to ramp back up, causing a fast over and underfrequency swinging, where the overfrequency swinging also led to wind turbines being shut down in the thousands, making the problem even worse
Power grid automation and protection are the industry I work in (R&D at Schweitzer Engineering Laboratories). This is one of the really pressing and interesting challenges which modern grid operators are dealing with. Wind and solar both add quite a few challenges when it comes to grid stability, but the economics are such that they are going to continue to become a larger piece of the puzzle moving forward, so we have to continue learning how to protect and regulate our evolving power grid. This was a very nicely produced video which strikes a good balance of talking about the challenges and reasons why power production still is going in this direction.
You work at SEL!! How cool
@@SeeNickView I think so. I love getting to work with the talented team of engineers I support, and really appreciate working for a company which has such strong values (and really tries to live them). I really appreciate how much emphasis there is on doing the best we can to always do right by our suppliers, employees, and customers. And getting to work on some really cool products and help solve fascinating puzzles is just icing on the cake.
Love your products.... The rtac software can use some work but the 351 series is my favorite relay, also really enjoy the 401 technology for fault detection and the new port servers.... Keep up the amazing work over there!
Now I have a name to curse at when I can’t get an SEL relay to test right (which is *definitely* not 99% user error)
@@nope653 I work a lot with 61850.
A dive into grid forming inverters would be an interesting future video.
I work at a Public Utility District along the Columbia River in the Pacific Northwest. We are also a Balancing Authority for WECC. There are many systems that are always being monitored but load and it's effect on the frequency impacted by that demand is closely correlated and adjusted by adding or removing generation capability on the grid.
Connecting it to the grid is easy.
Making it not explode when it does is the problem.
I think making anything not explode when it explodes is problematic.
Or being fried.
Anything renewable is a PITA to connect to the grid outside hydro because of the insane variability of the power output.
Or in other words:
"Everything can be connected to the grid at least once"
You can technically connect anything to the grid for a few milliseconds.
Another great video Grady. I've worked in the inverter industry for over 25 years and this video has taken all the complicated technology and simplified it to the point that a larger audience can gain knowledge. I'll be using this video as a supplement to the training material I've created. Well done sir!
I've worked in the inverter and battery world for 20+ years. This was a well produced overview of inverter technology
Matt, Greg from Alaska here. Hope things are going well. Talked to Lones every few years :-) maybe he’s the only old OB tech left. Anyway, take care of yourself,
@solarwind907 hey old friend! Good to hear from you. Lones retired and living island and boat life. Take good care. M
@@mattjames7272 if you run into Lones, give him my best,
I live off grid with outback radians good stuff
This helps me to appreciate what is going on inside my off grid solar system controller. Cooling fans spooling up and down in proportion to the amount of sun juice coming from the panels, amp meters on battery management units counting out the juice coming and going to keep me connected, comfortable, and conserving in an independent way. Super technology and we are gleaning the results of a massive effort. Thanks for being the one who explains the mega-systems under construction today.
Fun fact about the frequency. Microwave ovens usually use grid frequency to keep clock in time. That's why microwave ovens' clock never stays on time, at least very long. Where I live the clock ticks ever so slightly too fast, because our grid provides slightly over 50Hz power, but MWO still ticks on one second every 50 oscillations.
We have a 60hz grid and the operators actually average out the frequency by making sure there is 60hz*60sec*60min*24hours of cycles per day. Over time the number of cycles is about spot on, so our grid-based time keeping stays spot on over time, even if it varies a tiny bit throughout the day. Short of a power-outage throwing that all out the window, the only time I adjust the clocks is evil daylight-savings.
Yeah, the 50 Hz in Europe may vary a bit, but the frequency target is actually biased such that the summed cycle offset tends towards zero. There was an issue a couple years back when a dispute between Kosovo and Serbia resulted in underproduction and a prolonged lowered grid frequency. The noticeable effect of this was that frequency-tied clocks were running 6 minutes late. They regained this time over the following months though, as designed.
In my experience, there's 2 reasons a frequency-tied clock can get a permanent offset: 1) Your local area was islanded and the grid frequency was shifted to trip solar inverters. The change may be as much as 2%, resulting in clocks rapidly gaining time. 2) A simple blackout: No grid means no cycles.
@@AlexandervanGessel My microwave oven has been being running fast since we bought it over a year ago. It's not much, like 1-4 minutes a month, but never runs late.
I'm not sure if the Kosovo-Serbian situation affected us, we are in Nordpool and AFAIK we use almost exclusively our own electricity, and import rest from Sweden. I can be wrong 'tho
@@TheSanpletext Well, I'm referring to the central Europe synchronous grid (which stretches from Morocco through continental Europe to Turkey). While Norway, Sweden and Finland are members of ENTSO-E, they have their own separate synchronous grid, which includes parts of Denmark. Maybe the frequency is less controlled there?
Frequency reserve response (FRR) is the holdback. Hydro units shine in this area given they have the largest rotating mass. When an event is triggered the MW setpoint will be released and the governor will respond by either picking up or dropping load (the event can be over or under frequency).
I work in power generation ( a good portion of my work is in the renewable sector). Solar farms present a unique challenge in the fact that there is no inertia behind the power production. A regular power plant ( gas or steam turbine) has an effect almost of a synchronous condenser ( a small power reserve). Inverter based generation doesn’t provide that. Which can be weird once you start getting into crazy demand days.
Most inverters used on the grid are what’s called voltage source inverters, meaning that the dc voltage MUST always be higher than the inverter AC output voltage. This is to ensure the inverter can “pump” power out to the grid.
From a dispatch standpoint , most solar farms are not dispatched, they produce what they produce to the grid.
The ac voltage that gets exported to the grid is a relatively stable voltage regardless of the dc production, although the power flow will adjust with dc production. There is very little tolerances when it comes to grid tied systems.
Most solar inverters and battery storage inverters ( again utility scale) are the exact same equipment. They can be configured for each operation. Almost all inverters from the major manufacturers, are capable of grid forming, which means they can start with no grid reference. These units can be paralleled to create a backbone to which thermal generation can actually sync to.
Basically the more solar farms that pop up, the more battery storage systems will be needed.
All utility scale inverters have some sort or primary, secondary, and reserve frequency response algorithms built in. They’ll just convert this mismatch into heat ( vars).
Let me know if you have any questions. I can provide clear explanations on anything else. 😊
How big is the battery required for grid-forming? I'm guessing not by much, as it's essentially used as giant capacitor. Perhaps supercapacitor can be used instead?
@@hanifarroisimukhlis5989 they aren’t usually separate pieces of equipment. A battery storage facility has 2 electrical measurements.
Power: expressed in megawatts, which is the amount of real time power that can be produced.
Energy: expressed in MwH, or how long it can “generate”
The black start BESS systems I’m currently working on are in the range of 3-5 megawatt systems, with 20MwH of capacity. So it can discharge at full capacity (5MW) for 4 hours.
You need capacity and power for a black start to work, usually black starting the transmission grid takes time, even if successful the first attempt. You’re looking at a minimum of 3 hours to restoration depending on what caused the grid collapse
I was waiting to hear "spinning inertia". This whole subject came up when i did my degree back in 2017 and the exact predictions of a semi-blackout.
No grid tie inverter I have ever worked with works without "grid reference". Id est, without the 50hz reference point they cannot output. What century are you from?
I work in Scada for an EPC, those are good points. Would have liked him to dive a bit deeper. Seemed most of the video presented a challenge that none of our jobs usually face (the grid following). All of my jobs have PFR
Great job considering your world of Civil Engineering is quite a bit different wheelhouse than my world of Electrical Engineering. The depth that you dove into the EE aspects of power is very much appreciated by your entire audience, regardless of their profession.
tbh, modern engineering education (by 'engineering' here I mean application of physical sciences) should be changed in a way that it prepares students capable of understanding all forms of engineering as they proceed in their degree and could work in any one of them they choose, of course, they will undergo training as they start in the industry after finishing academia. They will have more clarity and higher intellect.
Great video! I work in utility solar and you did a great job explaining the complicated challenges we face.
I learned a lot about the power grid from this video. The graphics were super helpful to conceptualize. It was also telling to see the choppy waveform of cheap AC inverters. Better than a square wave, but not by much!
Excellent coverage of a very technical subject. I worked as a protection engineer for 42 years. Originally when small wind machines were being added to the system it was common to have them trip off for any perturbation on thw system. But as wind and solar grew to be a significant percentage of generation we began requiring ride through ability. Of course the vendors all had their own designs and everything was proprietary and often didn't react as expected by us.
Inverters also create issues for protection engineers as they don't produce fault current, they are current limited devices. That makes conventional relay protection unsuitable and more expensive high speed communication based schemes are needed.
IIRC fault-ride-through is actually designed to inject reactive power at possibly full kVA rating when a fault voltage pattern is detected, to trip the breakers without needing to have excessive real power on hand.
@@namibjDerEchte I work with 61850 monitoring - hardware and software.
Another exceptional piece of work, Grady. I’m proud to be a patron of your channel. My Father-in-love was head load dispatcher for PECO back in the day (retired before Y2K) and I’ve shared this installment with him. I am sure he will enjoy it. Blessings, my friend. Keep up the great work!
It's even harder when power companies like PECO don't know how to: read solar schematics, inspect a customer's transformer size accurately, or keep to their own scheduled appointments. We've had panels sitting on a customer roof for 4 months because of the incompetence (maybe even greed) of the power companies. Meanwhile, this customer could have been 100% off the grid and feeding excess back into the grid which desperately needs the extra power!
Well, you have to look at it from the perspective of the power company. Being forced to purchase electricity from every customer with a system is not a good deal for them in any way. Even if you set aside the fact that having numerous producers whcih you have very little control over complicates operating your system and makes it more expensive, just the purchase alone (separate from the operational complications) is a raw deal. The power company is required to purchase the electricity at rates based on their retail rates. Here's the thing though, those retail rates are a combination of generation costs and the power company's transmission and distribution costs. Building, operating and maintaining the infrastructure that delivers the electricity to the customer's home is a significant cost. 50% is a common figure, but I'm sure that it varies depending on the situation. Regardless, the price you pay at your meter is partly the cost to generate that electricity, and partly the cost to get that electricity from the generator to your house. Requiring the power company to purchase electricity you generate at that same (or similar) price means that they are paying way more than the generation cost for essentially additional generation capacity, while ignoring the cost to them to distribute "your" power to other customers. So given that, it's understandable that power companies aren't thrilled to be required to enter into agreements that just cost them money, and understandably aren't leaping immediately to the opportunity to lose more money.
How can one be 100% off the grid and yet feeding excess into the grid?
One of the better videos I have seen in a long time. You take a simple sounding project, like adding solar to the grid, and clearly explain many of the subtle problems that come up. I am going to use this video as a prime example of how simple projects can grow into complex problems. Thanks for all the hard work.
This is brilliant. I work for a large utility company (although I am in IT), and I love this video.
Talk about job security. 🎉
@@CheveraChino your comment make zero sense.
Many thanks to the people who have created this massive system, and who maintain it. I’m sitting comfortably in my air conditioned house watching RUclips videos on a Saturday morning thanks to all of you!!!
I found this super interesting! Usually you only hear about the problems with new technologies, especially when there's a powerful political lobby against them. So I was under the impression that renewables are an unqualified negative for grid stability and that we'd need flywheels or something if we get rid of too many turbine-based power plants. And I don't live in an anti-renewable bubble mind you!
So thank you for teaching me that these new types of power plants aren't better or worse, they're _different_ and we need to learn how to integrate them well into our existing infrastructure while it is changing at a very fast pace. This seems to be a big challenge to engineers and regulators alike, but one that does not seem to be beyond their abilities. Well, not beyond that of the engineers at least.
Fantastic Video. Grady... you really should add a link to the black start video in the description for this one! Maybe even at the end credits. There is a strong connection here.
It's amazing how much smarter the grid is becoming all the time. To think - a hundred and fifty years ago all we wanted was electric lights that could withstand a lot of variability and now we have all this delicate machinery in our homes that could be thrown off by a few volts at the wrong time.
Depends, most DC items use a switch mode power supply, which are very forgiving. They're rated (usually) 100-240v 50-60hz, so anything near or within that range will be just fine. The most sensitive common item will be things like compressors and other motors
Actually more of the "smart" devices are problematic for the grid during brownouts, not the other way around. Switching power supplies will try to draw the same amount of power regardless of the input voltage. Normally if the grid sags, motors and basic resistive loads will draw less power and it can stabilize, but too many adaptive loads will pull ever more current as the voltage drops.
You should watch the video again, we're not talking about voltage but frequency. It's frequency which keeps the balance of active power in check. Voltage regulation is less critical than frequency.
@@Chopper153 it's possible for a video to inspire other thoughts in people, you know? The delicate nature of our modern electronics contrasting with both the complexity of the grid and how much it has had to evolved since it was first being built.
Most of it shouldn't even require the power it does, but alas, the Internet of Things.
You really should research grid forming inverter technology. Every battery connection in Australia I work on now uses grid forming technology, which does not synchronise with the grid based on a PLL. Almost all projects in Australia are now also hybrid PV/wind and BESS, meaning many of these concerns are being addressed in all new projects.
I work in Utility Scale Renewable energy, and we also need to ensure grid relaibility for LVRT, HVRT, Harmonics, Frequency etc. This episode quite nicely shared the challenges of large scale renewables.
But aren't the big inverters (+100kW) doing all of that already on their own? At least the ones from SMA and Huawei do.
I hadn't heard of this event before but after watching the video, I'm pretty sure the power plant that experienced the failed arrester was one that I was the Lead Electrical Engineer for during is design and construction back in the early 2000's. I've spent over 25 years designing and building gas fired power plants as well as another 10 or so years developing utility scale PV facilities and this video does a wonderful job explaining a very complicated subject in a way that I think is understandable to most people. Keep up the great work.
I was camping just the other day and my friend who works for GE was telling me about synchronous condensers. Interesting that this video comes out the next day. Was hoping to hear a bit more in depth about how we manage, at least outside of inverters.
"Everything is a smoke machine if you operate it wrong enough."
From what I recall, in the early days of home solar. There was the issue of these small systems de stabilizing the relatively small Hawaiian grid that all home solar was ordered off the grid until things got figured out.
There was a problem with too many producers on the same local wire pushing the voltage too high, which was fixed by requiring inverters to stop feeding above a certain voltage. I don't think they orderered anything off the grid.
The joy of crating those videos certainly does come across, and seeing someone who's excited to talk about stuff they like makes watching that much better :)
Audio nerd here and you stole the words out of my mouth with that inverter explanation when I thought "That just sounds like the way pulse width modulation works."
Same here! I saw the stair-step waves and thought, "Digital audio!" I never realized inverters are basically 'digitizing' the power to make it into A/C.
Though normally solar inverters are horrid RF noise generators, as most of them really do not have enough filtering, and often enough the connection methods used are not the best from a RF noise point of view. You really need a lot more filtering to reduce harmonics, and such filters at power are both heavy, expensive and tend to run somewhat hot, losing power in the noise by dissipating as heat in the inductors and capacitors. No way to work it with smaller power, though you can reduce emitted noise quite a lot simply by connecting grounds correctly, and using properly RF bonded steel conduits for the cables, and a local RF functional ground as well. But most will not install that, as it adds a lot of cost to the system for little perceived improvement.
Reminded of my first time finding out what PWM is. Had to change the cooling fan motor on Lincoln and I thought it was broke. It wasn't. It was turning at whatever speed it needed to. Once I got it hot it went 100% full blast. Then I found out it was PWM and explained it to the owner, he thought it was broke because it wasn't turning as fast as he thought it should.
MPPT: Am I a joke to you?
@@cjc363636”stair stepped” waves do not exist in digital audio. That is a myth and misrepresentation.
Thanks for the excellent video as usual. This is a side of renewables that is seldom talked about outside of engineering circles so I am very happy to see such a good video on the topic. I have a clear memory of my 2nd year power systems class where we were shown a demonstration of how nicely "spinny things" stabilise the frequency when demand increases.
now that 2nd year class is probably tenth year :)
Great stuff!
Got a solar install here (in Ireland) which allows for off-grid usage during outages though there is some downtime in the switchover as it has to completely disconnect from the grid to operate. Still a useful option if you can get it.
How does the solar function after sunset? Do you have battery backup?
@@johntex105 Yep, though we haven't yet had a night time outage.
@@fuzzix Without having any clue about the amount of battery storage of the setup I reckon it could be quite useful. Should be sufficient to run the lights for a while and keep the fridge and freezer at the intended temperature until the sun is back ideally without clouds.
I remember a power outage at my parents once during the winter, one complication of that with central heating I had never thought about at that age was that it'd kill that too. Sure the actual heat comes from burning gas but without electricity to pump the warm water around and run some electronics involved with the burner it'll just sit there being pretty. Fortunately it wasn't all that long and they have a wood fireplace in the living room, but older buildings sure can get cold quickly if it's freezing along with a strong wind.
By the way, the reason why they're called 'inverters'. Back in the old days of mercury arc rectifiers, there was only one type of conversion you could do to electrical power: AC to DC. Any unit that did AC to DC was called a 'converter'. Eventually, technology moved on and it became possible to go from DC to AC, the opposite process of what a 'converter' did. Hence, it was called an 'inverter'. Since then, the term 'rectifier' has become common for AC to DC ('PFC' appears occasionally, don't get me started), but 'inverter' for DC to AC has stuck.
It sure is difficult. But engineering is transformative-no matter what field or area you're in.
Nicely said.
Yeah nice one, morning after Barbie
I'm a consultant electrical engineer for many large US utility companies. There is no plan at any level by anyone for a zero carbon electrical grid.
I absolutely love how clearly you discuss the challenges with inverter-based resources. I work in nuclear, and have had many discussions about how so many people think we can run the whole grid off wind and solar, but there are so many challenges that spinning turbines help resolve and inverter-based resources would struggle with. But discussions so often devolve into mud slinging about the demerits of each generation type rather than a smart discussion of actual problems and complexities of running a power grid. So I appreciate how you just discussed how it works and is operated in this video!
In Toronto, you can either connect it to a battery, or the grid, but not both. They won't let you use an automatic switch so that you can sell your solar power when the grid is working, but then use your solar power during a blackout.
You get grid tied inverters with battery now, that will charge the battery as priority with solar input, and excess will be fed back via grid tie, and after dark they will run off battery to a point, then leave the battery and run off grid till morning, when they use the solar to charge the battery again. Manage your power use and you really can be a net exporter, never actually drawing power in from the grid.
Utilities increasingly make it difficult to sell power back for any meaningful gain. Solar panels can still make sense, but one really has to run the numbers carefully. Assume worst case scenario for uptime and maintenance. Panels can last decades, but many components won't and are consumables that need to be periodically replaced. To put it another way, figure whatever a solar panel company is promising is fantasy and budget accordingly. As in 10 years payback at most. If it's longer than that, likely best to wait it out for now.
@@ronbennett7885 There are a couple of issues here. First, often solar panels are producing power when there is excess in the system, so it not valuable. But if folks do get high feed-in tariffs or net metering schemes, you essential have poorer people paying for rich folks having solar systems. Third, people scream bloody-murder about having to pay for grid connection fees when they are using little or no power. But someone needs to pay for the grid.
@@ronbennett7885 oh back when I did this, the Ontario government was buying it back from us at 55c/KWh. On a 20 year contract. It's still going! Just can't actually USE the solar power cause the electric company doesn't trust the auto switches and doesn't want me sparking their sparkies.
@@richdobbs6595 The high tarrifs were to encourage investments back when solar cost a lot more. Now new investments do not need the high tarrifs.
It seems a bit like the answer in the Edison vs Tesla debate may be shifting.
The decision for AC came from a time when phase regulating equipment was only needed in a hundred or so power plants across the country while inverters were expensive mechanical devices, but it allowed us to have "dumb" transformers in hundreds of thousands if not millions of places across the country. However, now this means that we need "more modern" phase regulating equipment with the inverter in each solar, wind and battery installation. And with our goal of expanding renewables, we'll likely see solar installations on the roofs of a sizeable fraction of houses, and thus possibly more inverter based sources than transformers. I wonder if there would ever come a point where it would be cheaper to have built a DC grid instead.
We'd need an own inverter on the input side of each transformer, but they just have to be the solid state kind without any synchronization features that make the "more modern" ones more suitable to the current complex landscape, since they can be independent and don't have to "play along nicely" with the rest of the grid. As far as I understand it, the fact that the frequency gets messed up whenever there's a disturbance results from the fact that there's more than one source of AC power on the same wire. With a DC grid, that can entirely be avoided and transformers should work at any input voltage (they could merely output AC power with a lower voltage, while the frequency can be held sufficiently stable to protect appliances).
And even if the DC path never gets cheaper, the AC decision was made at a time when only factories and street lighting really relied on the availability of electricity (if they weren't still gas lamps). These days so much more goes down when electricity is out. Not just all indoor lighting, communications, all of our home appliances, and hospitals; but also our heat, water supply, traffic lights, shops, and the power plants themselves. (Yes, there's exceptions whenever outages are too disturbing that backup generators or USPs are kept on standby - which also costs money.) The resilience of not having a single bad law, guideline, or default setting in a series of inverters turn a slight disturbance into a full power outage, might even justify a solution which costs more but isn't so complex that predicting it's behavior becomes near impossible.
I mean, even running the self-regulation on the current power grid requires a separate data channel to communicate the current (and possibly even forecast) electricity price. A DC system would in theory regulate itself to some voltage value in the acceptable range, even if you did something as simple as setting a convention how the electricity price varies with grid voltage, possibly even with additional dependence on the time of day to further stabilize the voltage against predictable cycles of demand (and supply).
That moment actually passed a long time ago - all computing runs on DC (transistors are DC devices) so we have to convert AC to DC for every single device which uses a computer, chip or logic circuit, which is everything in (not even modern) electronics. The cumulative costs of that passed the costs of a DC grid ages ago.
But there's never been a point, and never will be, when making an entirely new DC grid and switching everything over from AC to DC is cheaper than just continuing to make do with the AC grid we have. Not only do you have to build a DC grid, but you have to rewire every building and replace every device built for AC. Even all the devices that would work better on DC would need to be replaced because they've been built for AC input.
The cost would literally be all capital investment across the entire economy for several decades (that's what you're replacing), just to get to the same point we're at today with a DC rather than AC grid.
If only Tesla and Edison had a crystal ball that predicted the technical workings of computers - they could have made the right choice, and also invented computers.
I know of a few new factories here in Austria that use two separate power grids. One classical AC, the second one DC to use the electricity from the solar panels on the roof directly.
Where are they located, which companies?
It's interesting that, DC-sources are causing people to start thinking outside the AC box, and start looking at using the DC directly (pun intended). The simplification it creates has benefits.
I've wondered when DC end to end would start happening (even if in small scale), given that more and more loads either use DC (lighting, electronics), or in principle could just as well use DC or AC, while there's increasing generation that's inherently DC (solar). As an aside, AC is responsible for much loss in the grid (for the same reason that transformers work... induction). More and more long distance transmission lines are DC. So the chain is sometimes DC generation, DC for part of the transmission, and DC loads. I've read that some companies are working on fully DC grid designs. I doubt that it'll become the norm in my lifetime, but I wonder when the benefit of transformers won't be sufficient to justify AC, if DC grid technology gets sufficiently good someday.
@@bearcubdaycare one of the problems is, DC doesn't work well over longer distances, not without some seriously big cables. but if you were going to do certain dedicated things, like say a direct PV connection to an electric hot water heater, then it might not matter as much.
It tickles me to think of fancy new Austrian factories working on basically the same principle as the old camper trailer I used to have (which had an external AC connection for the aircon, appliances, and wall outlets, but ran all the lighting off an internal 12V DC circuit powered by a car battery). ;)
I am not an engineer, but I have a personal interest in such things and have picked up a good amount over the years. That said, some of this was slightly over my head, but you present in such a way that an enthusiastic amateur can gain a greater overall understanding of the situation and the problems that are trying to be solved. *Thank you* for your efforts!
I've also heard about the problem of voltage rise near solar installations- as solar panels doesn't use the concept of power angle like generators, the only way they can inject energy into the grid is by increasing the voltage a little bit. This is fine for a few panels, but in residential areas having large number of installations, each panel has to up their voltage, resulting in significant rise in voltage in lines in the vicinity.
Mostly because the grid was designed as one way, you would need to have a lot more automatic on line tap changers in the residential suburbs to counter this, as they typically are only used in the larger local distribution network for the primary incoming transformers. Local uses mostly off line tap changers, set, locked and left as you cannot change them energised. Automatic tap changers in areas with lots of solar will help, but the expense for the utility is not warranted, seeing as it is not power they are able to bill for easily, and thus low on the priority list to implement. Residential inverters are programmed not to exceed peak line voltage, so the power loss is purely on the solar household side, not the utility side.
Usually there are certain limits to this effect, so that PV inverters will lower power to prevent the voltage from rising too much. In addition, there are actually concepts to counter the effect by drawing reactive power (lowering the cos(phi) of the current injected in the grid) at the same time. I don't know if they are implementet in current PV systems, though.
I dunno about "significant", but your surmise is wrong... grid-tied inverters do use power angle like generators. They push phase, not voltage. But that does have two side-effects which do cause the voltage to rise.
(1) The homes are burning less energy from the grid, the lighter loads mean higher voltages.
(2) The exported energy has to go somewhere and since it isn't going to frequency that leaves only voltage. Even though the inverters are pushing phase, there will be some voltage rise on the whole circuit as the energy smooths out over distance.
But again, "significant" ? Its possible but exceedingly rare, and it will still be in-spec since the grid-tied inverters trip-off if the voltage goes out of spec.
A very real problem is that voltage rises in residential voltages feed-backwards through the pole AND substation transformers which multiply it back up to transmission voltage levels. And even when these rises are within specifications, the multiplication on the way back, if the backfeed gets all the way to the substation, can cause the transmission line voltage to go out of spec.
This is easily rectified but requires a transformer upgrade at the substation (basically to a modern transformer with two-way sensing and automatic tap control). Fortunately, though, this sort of situation only occurs with a very small number of circuits, so mostly problems develop because utilities are blowing smoke and purposefully letting them get out of control to try to convince politicians to slow solar installations down.
-Matt
Thanks for the simplicity in explaining this topic, I did my Ph.D. in this topic and they way you explain is remarkable
All that spinning mass is the stored energy to get through transient events. On an airplane I worked on there was a motor generator to both get a particular DC voltage an aftermarket system needed AND it also ensured clean, stable output even if the input AC was interrupted briefly.
An NC (Asheville?) TV station was said to have had a flywheel 'buffer' between the mains grid and the on-site diesel generator. That way, when grid mains were lost, the switchover was not a blackout / brownout event. The spinning flywheel buffed out the transition. Sorry, but I don't remember the call letters. I'd heard this from TV engineers back in the 80s/90s.
I used to work for a Utility level inverter manufacturer and our California customers often opted for the Ride Though board. They allowed the units to keep running for a couple minutes while the grid went through a black out (happened often back then). Fun times!
5:31 I audibly laughed when those skwigli lines poped up. Is there a schematic of that cursed thing by any chance?
Should send it to Big Clive to tear apart and show how it (barely) works!
I wanted to watch this because my favorite presentation at a data conference a few years ago was from an electrical engineer from Dominion Power (my power company). He started with a bit of an overview of electricity and went on to show how his team was monitoring fluctuations in order to enable the power company to better handle the increasing domestic solar installations. But you covered a different angle! Interesting information!
also, i loved that talk so much that i asked several other ppl if they liked it as much - they all couldnt follow his intro overview at all - I guess the 2 years I spent as an EE major were good for something.
Afaik in Germany you are required according to vde-ar-n-4105-2018 "Therefore, in the future, newly constructed generating plants must support the grid in case of disruptions."
That would mean more complex and expensive power conversion equipment, but it's what's needed so it can come out of the corner as more than a niche player.
@@SeekingTheLoveThatGodMeans7648 Ehhh, not really: at worst you need a small battery about 20% as expensive as the solar panels themselves, and that's for a much higher level of capability than the regulation mandates. Those mandates are basically all just controller/brain requirements, which may have some effect on what sensing circuits the inverter needs to feel the grid, but nothing that actually handles power.
That's the beauty: the normal inverter already needs to have all that power conversion circuity just to feed a reasonable sine into the grid when the sun shines.
I work on photovoltaics and batteries directly, but it's really cool to see the other side, on how people work on installation and application. Thanks for this!
the other reason home solar arrays typically shut down is because they are not allowed to remain connected to the grid in a power outage, and it's easier to shut down the array, than to have able to disconnect from the grid and operate as a standalone system.
Especially if the inverter connects in parallel with the grid, it's impossible for it to isolate itself. A smarter inverter MIGHT be able to do so.
It's a function of money. This technology already exists, it's just not practical for an individual homeowner
@@thedude5040 a standalone system pretty much has to have a sizable battery bank and be able to manage production based on the battery bank. and have an air gap transfer switch to disconnect from the grid. doable, but people haven't started asking for it in enough numbers to inspire manufacturers to go there.
@kenbrown2808 why invest $30-50k for this kind of setup when a $600 gasoline generator with a 5 gallon gasoline can does the same thing.
@@thedude5040 it's not going to cost that much MORE than the grid connected system. and a 5 gallon gasoline can is good for about 8 hours, depending on the generator.
It’s not just a challenge to provide ‘fault ride-through’ during transients, but also to provide sufficient fault current at the source, for the generators etc to trip correctly without damaging themselves. If there’s insufficient current available at the generator, as is often the case with renewables (they don’t individually have the inertia of much larger generators), especially when connected over long distances - the protection won’t function correctly and they’ll end up burning out equipment and cables as they unable to ‘disconnect’ themselves. It’s why many grid connections of renewables require large synchronous condensers or similar, to provide the immediate energy/inertia required for protection at the generator source to operate quickly and correctly.
I didn't understand why grid frequency drops when there is more demand than load, but this makes sense:
>whenever there is increase in active power requirement, kinetic energy of rotor is used to supply this increased power demand which lowers the speed of rotation and results in frequency drop.
THIS‼️ I am really confused as to why Grady did not discuss where the added power comes from. My assumption is the same as yours, but I did not hear it in the video. And, I am shocked that I had to scroll through 100s of comments to find a mention of this. I feel like I must've missed something.
Practical Engineering prepared a good video on the subject. The take aways are that with the present technologies the grid will need conventional rotating machines to maintain the frequency and the voltage support for the frequency following inverters to function. Following the technical literature there seems to be a limit for renewable generation with frequency following invertors at 70% penetration due to dynamic characteristics of the grid following a major disturbance like a fault. As cited in the video there is a need for inertia and the ability for reactive support from generating sources to ride through the disturbance. One must remember that the invertor-based devices can provide some of these requirements, but there are limits within the controls for them to be effective. For the grid operators the information of those limits needs to be well understood. There is plenty of work to be done for achieving the goals.
oh man, you gotta make sure the magic smoke stays INSIDE the equipment!
Smoke is a process indicator; fire is a fail.
@@RonTodd-gb1eo Cool! I'll have to remember that!
In Finland, it took over a decade to make necessary modifications to the grid, and we are finally in expansion upgrades only mode. And we had a state of art power grid already in the 90s.
You nailed this subject. I will be sending this video anytime I have to explain how grid frequency stability works.
7:02 Awesome explanation. I've always wondered why many grid tied solar kits were useless during power outages.
An enlightening video. Really shines a light on the topic in a way any sun of a gun can understand.
commendable civil eng. channel, deep dived into electrical/electronics, likely required consultants for this video. however, it was only difficult for a brief period when renewable energy is rising and battery energy storage isn't common affordable yet. these days, this problem is virtually eliminated by BESS and smart grid systems.
Just started to explore Grid-tired solar system. This video came a a good time. I was discussing with our Solar vendor about having the inverter continue powering thru a grid outages... and the Solis smart inverter could not do it. Yup, knew of MPPT, and how Solar Generation is not as smooth as most people think that it is, passing clouds could drop the generation just like that, and it will cycle thru generation at Peak and trough...not an easy task, trying to do power generation, and still have to follow the grid's frequency and voltages.
What ever you choose, choose hybrid inverter with un-balanced output with some batterys.
With a grid tied inverter, increasing or decreasing the voltage will just push or pull VARs from the network. It is actually shifting the phase of the inverter ahead of the network that will push power out.
It's interesting to think about how solar and distributed inverters could be used to provide half-cycle ride through and power factor adjustments. Seems like just changing our thinking we could use these to our advantage.
you mean from profit based thinking to solution based thinking?
@@kenbrown2808 One does not exclude the other. You provide a solution and make a profit. That's capitalistic 101.
@@huckleberryfinn6578 no, that's economics 101. capitalism 101 is "a dollar in the hand is better than two in the future"
@@huckleberryfinn6578 That's Adam Smith air-quote "capitalism". It relies on this underlying assumption that the best solution will always be the most profitable. It doesn't take a Nobel laureate to figure out that assumption just is not the case.
PG&E is a classic example. Much of their infrastructure is decades past its rated service life. But upgrading and replacing it cuts into profit, so instead they cut power to their customers and burn down the countryside whenever there's a light breeze.
If you care to educate yourself on some of the thicker weeds, you should look up concepts such as price elasticity of demand and hydraulic despotism. These two principles are the Achilles heel of capitalist ideology.
This was mentioned, just briefly, in the video - some inverters can provide what I think he called 'virtual inertia'. Either boosting the front of the sine wave or loading it down if the frequency drops or peaks.
I remember that balloon escape from East Germany back in 1979. That was some pretty creative engineering just to get it to fly in the first place. But then to actually make it across the boarder even with multiple mishaps along the way, is amazing. As a student in 1972, I got to spend a summer behind the "iron curtain". Those governments didn't fool around, there were police everywhere.
Another example of grid frequency blackout is the eastern seaboard blackout of 2003. The grid was running at capacity when a powerplant tripped, and 8 states and southern Ontario tripped offline.
Power lines have their own internal resistance (called volt/amp reactance or VARs), and resistance converts current to heat. The more current you try to push down a line, the more heat created and the line starts to expand and sag.
You can actually see this if you photograph a big power line from the same spot during high and low power demand periods.
In the case of this blackout, a line heated up and sagged into a group of trees that hadn’t been trimmed. This caused a ground fault, and a protective relay at the plant did what it was supposed to do and opened the generator breaker, taking the plant off line.
This caused a drop in frequency at nearby powerplants forcing their under frequency protective relays to open their respective generator breakers.
This lead to more and more areas where grid frequency collapsed, causing a cascade failure that took out most of the northeast grid before it could be contained.
The next problem was how to bring the grid back up, and it was caused by people. They wanted to know when power came back, so they went around turning on lights and radios to alert them when power came back up, creating a huge potential load.
Anytime a powerplant tried to close its breaker, it would just get dragged back off line. Since the grid was dumb (no remotely operated switches) in 2003, (most of it still is dumb), utility workers had to go out and manually open switches to “island out” areas small enough that the generator governors could react fast enough to the sudden load without tripping.
At that point, the people with power would go around shutting off all the lights and radios, lowering the load in that island. Then the next island could be connected in and so on. It took 7 hours to reestablish power to most customers. It was several days for some customers.
At no point did any equipment fail. It all did exactly what it was designed to do. The point of failure was the lack of maintenance in trimming trees (exacerbated by environmentalists trying to prevent tree trimming).
The grid operators in Ohio were criticized for a lack of understanding of their section of the grid and improper response, but in fact, understanding grid operations requires being able to do the calculus in your head. A bunch of business majors had no chance.
I was stuck at JFK airport until next day for a flight to Europe. Slept on a baggage conveyor. Nice music from a traveling orchestra.
That was really informative. Thanks. I’ve played around with a couple of 400 w solar panels and have run across a lot of terminology that I did not really understand till now. Funny how a small scale solar setup can help make these huge installations easier to understand
Ah, I was wondering why we had a power cut... when we had a power cut. Figured there was something about our solar configuration that meant it couldn't work as the sole power source for the house.
Grady did not cover another reason solar systems turn off when there is no utility. If your solar system is feeding power into a utility that the power company has turned off for repairs then you are energising cable plant that is expected to be de-energised. This can be dangous to linemen. Remember transformers work both directions.
There are three kinds of solar systems. Off grid, make power all the time, can't be connected to the grid cause it doesn't sync. Grid tie, which does sync to the grid, but when the grid shuts off, it MUST shut off, both because lineman don't want to get zapped, and because you will never own equipment powerful enough to carry the whole neighborhood. The third is a hybrid, which can do both, because it automatically disconnects from the grid, and then turns on(often with multiple pieces of equipment). Tesla has a system. They used to have a Tesla gateway, which was the switch that disconnected you from the grid, and let the powerwall know it can turn on. They might have built that functionality into the Powerwall 3, not sure, but their new setup isn't approved everywhere yet. Enphase has a system. SolArk sells a hybrid inverter. Point is, if you want your system to do it all, you have to design it that way, and pay for the parts. I just have a grid tie from 2017, but I want to eventually be able to do it all.
Your videos on grid and power are my favorites. I practically drop everything to watch them. Love your book too!
its practical to drop things? ok if not carrying a cup of cofee i suppose ;)
Solar is excellent for future-built homes. You add DC outlets (or USB ones) designed specifically for electronic devices, power tool battery chargers, or electric cars so you can directly power them without having to invert it to AC and back again.
I think this would work well. Look at RV's, the have 12 volt, 120 volt, and gas systems, why not our home and work places. lights and electronics could all be DC.
I like the concept in that most electronic devices, if you look at the little wall adapters for them, run off of 5VDC or 12VDC. Its a whole lot more efficient to buck-down a modest DC voltage then to convert to AC and then back to DC. But there are some warts.
The biggest wart is a safety concern. You can't use AC breakers with DC voltages and you also have a problem with ARCing when wires break that you don't have with AC (because with AC the current crosses zero 120 times a second). So even though we can have in-house DC wiring, its a bit more dangerous when faults develop as the wiring ages.
As such, I think DC circuits are an interesting concept, but amperages would have to be severely limited and electronically controlled., possibly with an insulation resistance check as part of the safety device requirement.
I think something like .... well, for house wiring you don't really want 12V. You want something like POE+ (power over ethernet) which is an actively negotiated protocol which also actively checks available wire pairs continuously. It runs at roughly 56VDC. Almost impossible for POE+ to cause any sort of fire. It is typically limited to roughly 60W over four pair which is barely 15W per wire pair (0.35A or so per wire pair). There is also POE++ which supports higher wattages by using higher voltages.
-Matt
I'm still waiting on reasonably priced DC breakers to actually use that (not at my home, because I don't have a good place to put solar), particularly actually for a mini-split AC/heat pump.
@@junkerzn7312 Modern AC breakers actually have substantial DC capacity (and rating!), it's just that you can't really handle more than 60V with a single contact MCB.
There are low voltage DC standards in place currently for e.g. iirc a bipolar +-125V system without locking connectors, and an iirc about 380V system with locking connectors that's been deployed in data centers for a while now (their DC is their battery backup bank; they skip the inverter of their UPS).
@@namibjDerEchte Well, I've got like four different brands of DC MCBs in front of me and the unpolarized ones are typically rated to around 110VDC and the polarized ones up to roughly 250VDC, I think I have one PUFA here somewhere with a 1000VDC rating.
Generally speaking, the DC MCBs with higher ratings tend to be polarized. The DC MCBs with lower ratings tend not to be. But this is an area that get DIYers into trouble all the time. Either they buy polarzied breakers and don't hook them up properly, or they buy polarized breakers and don't realize that they can't be used in a battery circuit where current can flow both ways (charging and discharging). OR they buy a non-name breaker and just assume it is polarized or unpolarized from the markings, not realizing that most polarity markings are completely meaningless as a "tell" on the type of breaker.
Most AC breakers do not have DC ratings and definitely cannot be used in DC systems. Not sure where that came from. Expensive high-end breakers will often have both AC and DC ratings but most run-of-the-mill breakers that people buy... those almost never do. The construction is too different.
Typical AC breakers use relatively large GAPs to break ARCs for example. And also often don't have arc extinguishers. But that method relies on the current crossing 0A 120 times a second (AC waveform). DC with a breaker like that will build a lot more carbon up on every trip until the breaker fails and fails to break the ARC, then catch on fire,
Typical DC MCBs use narrow gaps and walk the current up to the ARC extinguisher. That can be done both polarized and unpolarized. Polarized breakers can walk the current more quickly and more reliably and tend to have higher voltage ratings. Unpolarized breakers still walk the current away from the contacts but don't have magnets near the contact point forcing it to walk faster. But being unpolarized, they can break current going in either direction.
-Matt
Actually most wind turbines usa an asynchronous induction generator. They drive this with a fixed gear ratio to maintain rpm just above syncronous speed. Thats why they always turn at the same rate, no matter how strong the wind is. When the wind is too slow to generate any power at this speed they turn all the way off.
You're describing a type 1 turbine, and they are much less common today compared to types 3 and 4.
1900s: "The future has flying cars!"
2024: "We _almost_ have, a power grid... still working out the bugs..."
To be fair, it's just as well we didn't get flying cars. A fair percentage of the population really shouldn't even be trusted with the normal kind.
Way oversimplified theory vs. real engineering that has to be tested to the utmost.
We have a real power grid
Helicopters are just flying cars, so.
@@Duiker36 I wouldn't like to try to fly one using my car-driving knowledge. ;-)
Let me just say that I am a big fan of your videos as I am an electrical engineer and I love your deep dives about practical engineering. I think overall, you summarized the topic of challenges regarding inverter-based resources quite well because lack of inertia, ride-through requirements, and protection issues are among the the top 3 challenges for sure. I am only commenting here because this is the only topic you covered that I have some expertise with. I really do not want to nitpick here but I felt like there were some small errors with the way you explained how grid-tied inverters work because once synchronized with the grid, they do not vary their voltage to control how much power is flowing into the grid because the grid will dictate a constant voltage and frequency. They vary the phase angle between the inverter voltage and the grid voltage to control the current flow and thus the power flow to the grid. The voltage amplitude stays constant and that is why the graphic is also misleading showing the voltage waveform going up and down while this is inaccurate. The explanation of how inverters work was a bit oversimplified for my taste of course and you showed a graphic of multi-level inverters instead of pulse width modulation which looks a bit different but I think that is okay since you are explaining the concept to a general audience.
Again, I was so excited that you finally made a video about a topic that I care about and know a bit more about haha. I am not pointing this out to criticize your efforts but because I feel like it is important to try to give accurate information even though it is challenging when you try to simplify it to be consumable to a general audience. Keep up the good work :)
Maybe someday we'll have cheap thyristors that allow abundant high voltage DC transmission
Highly unlikely. We've reached the end of Moore's law, so silicon fabrication technology isn't going to get much cheaper.
@@Shaker626 , grid level semiconductors are not using shrinking to improve economies of scale. They're much larger due to the power they're handling, so Moore's law has little impact either way.
@@JimBob1937 The wafer production backend is a big chunk of the cost of those chips (their fabrication is not as expensive, however). Moore's law was the main driver for making wafers cheaper, which is not very likely to proceed at the same rate. It's unlikely that power transistors will get much cheaper than they are now with the IGBT.
@@Shaker626 Moore's law is for scaling down transistor, not efficiency/power density.
@@Shaker626 , wafer production is based on growing a cylinder from a seed crystal and then slicing and polishing it up into a set of wafers. This has no connection to Moore's Law. Moore's Law is more about shrinking down the transistors via smaller and smaller interconnect technology and better and better photolithography (most recently EUV, due to the shorter frequency giving better detail resolve). This in turn ither allowed for an increase in transistors per fixed area of the wafer, or for less wafer area for a fixed number of transistors. This allowed the wafers to be better utilized. However, grid level semiconductors are not concerned with size as much as power handling. Nothing of Moore's law directly affects them, since it isn't an area they're concerned with.
4:37 .. These are still popular today ..
There are several, single to three, "rotary phase converter" companies, and the old "Redline" 12 volt DC to 20 volt AC "rotary converter" is still used today.
Grady, I'm sorry to do this to you, but it kills me to see smart people doing this. At 2:18, it's either "comprising" or "composed of." "Comprised of" is not a proper usage of the word. Just a PSA! Big fan though and I love this topic. These videos where you go over events are my absolute favorite. You're a clear, concise storyteller and super great.
This was a great episode. The comments were as informative as the content. The person from the EU who have enough solar in neighborhoods that when the grid trip off, the neighborhood may not due to backfeeding was something I never even thought about.
My only real comment about it was There Were No Googly Eyes On The Demo. There are always eyes on the parts of the demo. I would have thought the ideal place was on the hard hat sitting on top of the scope, so it could watch that perfect sign wave from the inverter.
Thanks for another great episode.
So consumers front the bill on infrastructure but energy companies absorb all profits
It's the American way, maaan. Like you wanna go all socialist on us? Ewww.....and so on and so on....sigh. Thank you Gordon Gekko, you've f***ed the country good and proper..
I know right? Why TF are all the power companies saying they need public money to upgrade the grid for EVs when they stand to sell two or three times as much power in the end?
I'm jealous of your ability to vulgarize these concepts to a larger audience, grid forming is so hyped right now ! These backyard demonstrations are also very cool :)
Can someone explain what he means by "I will not be making an interconnection here" at 6:41 and why the inverter would let out the magic smoke?
1. Attaching any (heavy) power equipment to the grid (i.e. an interconnection) requires permit from power utilities.
2. A badly planned interconnection will cause equipment and possibly grid failure, ending up "in smoke".
@@Speak4Yourself2 Thank you. I still don't understand why he calls it "making an interconnection" though, like how can he create one by plugging in a device?
In his demonstration, he showed the output voltage from his inverter and the voltage from the grid. You can see that the voltages on the screen are very unsynchronized. Practically, when power plants connect their generators to the grid, they have to adjust generators' voltage and frequency to be synchronized with the grid. On the screen, the frequencies of both signals aren't equal as you can see that the signal moves relatively to the other signal. And voltages are even worse. One is sine wave while the other is step. If you connect them together the difference of the voltages will cause high amount of current and thus cause the smoke.
@@BBP_BKK "Connect them together" as in connect the inverter to the grid? Why would that be a problem exactly?
@@sebastianelytron8450 Yes, connect the inverter's output to the grid. As you can see on the screen, the inverter tries to make its output waveform to be kinda step waveform while the grid is sine waveform. Two outputs with different voltages should never be connected because those devices want to regulate their outputs to be what it's designed for. (This applys to everything, USB, HDMI, etc. You have to connect from one device's output port to another device's input port.) Two outputs trying to make difference voltages will cause high current as the interconnected wire resistance are very low and I=V/R. It's basically a short circuit. I think you should check out the video titled "What Is A Black Start Of The Power Grid?" on the channel starting at around minute 10:00 to understand the term "synchronization" and come back later if you still have question.
Outstanding description and visualizations for those of us who are not engineers. I'm a laymen selling radio communications services into utilities and this kind of information is soooo good!! Keep it up.
It's always crazy to me just how much Texa's grid is privatized and how often that causes issues.
Great video, however there's a slight error at 6:14. You mention pulse width modulation with incorrect context: PWM has to do with the ratio of "on time" to "off time", most commonly used for pure square waves, and doesn't have anything to do with the number of steps used to approximate a sine wave. The term you're looking for is "switching frequency", analogous to the number of bits in a DAC.
We might need a video on the FSK Bridge
I built a solar powered kayak by putting a large flexible solar panel on a PVC adjustable frame. It acts as a sunshade, and powers a standard 12 volt trolling motor via PWM to provide the optimum panel loading. The fun part is this is done manually to find MPPT, so changing course or a cloud coming over requires PWM adjustment. Like sailing, where you need to constantly adjust for the wind.
Every South African hearing load and shed getting a PTSD response 😂
Bro 😂
The power conversion..... invokes memories of selecting Uninterruptable Power Supply units for some computers many years ago. Cheaper models put out a "stepped sine wave" and better units put out real sine waves. One section in the video explained details that I never knew (all I knew for UPS selection is that real sine waves were better). Thanks!
Some of it is just for safety, but 90% of the problem with solar in my state is political - the utilities have complete control of the legislature and the laws are punishing for anyone who isn’t them
This gives you an opportunity. You can get retired panels and connect them to an off grid load. In Phoenix I would put retired solar on the roof and connect it to a mini split for free air-conditioning
@@TimHayward Great use of solar.
For profit utility grids are terribly inefficient, from a customer value perspective.
@@TimHaywardRetired panels? Never heard of it, but now I'm definitely curious!
Why do they Retire them? Do they need some type of maintenance?
Where do I find these?
Thanks for sharing your ideas, thoughts and videos. As a retired senior system operator I always find your videos interesting and usually spot on. About the only difference is your use of scientific units. In the bulk power industry we usually talk in terms of megawatts for communication clarity. 500KW would be 0.5MW. Or my import limit would be 1200MW not 1.2GW. I have always enjoyed when people build simulations of larger systems, really liked your demo. Was reminiscent of a simulation I once saw using WW2 era aircraft motor generators. By removing the shielding you could see and manipulate the rotors. By tying them together you could simulate how the emf force thru the wires tied them together. Since they were permanent magnet motors. Hand turning the rotor on one generator would cause the other generator only connected by wires to turn as if by magic.
Another technology you may find interesting to look into are VFT aka variable frequency transformers. They are made by modifying a three phase motor with the stator windings as one half of the transformer and the rotor windings as the other. By controlling the motor input shaft with a stepper motor you can vary the phase between the two systems and push power one way or the other dynamically. An example of this can be found at the cedars station between the NY and HQ power systems. While learning of the technology when they were training us for its introduction to the power grid a fellow operator recognized he had seen a similar setup earlier in his career at a smelting plant. They used a hydraulic cylinder to push and pull on a crank arm attached to a motor input and rectifiers on the output. Making an industrial sized variable DC power supply.
Our south african eskom would shock you...
😂yeah totally....