TMC is an independent, primarily volunteer organization that relies on ad revenue to cover its operating costs. Please consider whitelisting TMC on your ad blocker or making a Paypal contribution here:

Power Factor/kvar/waveform marketplace

Discussion in 'Tesla Energy' started by bwa, Jan 12, 2016.

  1. bwa

    bwa Member

    Dec 8, 2014
    Aptos, Ca
    Reading about the canary cackle over in that Hawaiian island about his "frequency" problems with such big solar farms made me realize the grid marketplace is very behind in understanding the relationship between generators and users.

    The generators set the frequencies, waveforms, etc. (Failure to do so: the band aid that the generators are using other generators to piggyback this setting is ridiculous and temporary. Cryptographically authenticated synchronous secure communication of this needs to be distributed to all the generators, and that means grandma with her roof, and that bozo down the street with a zillion whatevers who calls himself a utility because of historical civil placement.)

    The users use said stuff.

    The generators make offers on the market for their goods.

    The users buy said goods.

    The roadways to and from market is the grid.

    In the case of alternating current (AC), apparently there's a bunch of users that don't use the AC waveform energy identical to how it's initially offered in a perfect sinusoidal form. I knew about this for ages, but recently (a few days ago) read about it as a discussion of Power Factor, kvar, a 90 degree out of phase load, capacitance, inductance, etc.. Whatever it's called, the key to me is that the solution was still as simple as it ever was: get the inverter to invert the proper waveform.

    In this environment, what that means is that every waveform, wavelet, wave piece, equation of wave, etc., is a product to be sold at market, offered for sale, negotiated, and sold, of course in continual periodic equational format, responsive to the actual instant demand. Grandma's solar panels (and by extension all utility generation) will offer its own waveform response characteristics (with amps, volts, etc., all in tow), as an item to be sold, at the point of sale (grandma's grid connection point, in her case). The considered purchaser (usually the grid operator as an arbitrage broker assimilating instant waveforms across a geographical area closest to their needs, but could really be anyone) then decides if they want to buy at that pricepoint (as a component of their overall use, of course), and then makes the decision. Grandma either places on the grid this energy or doesn't. If ordered, grandma offers the energy in the specified contract waveform. The contract may last for less than a single cycle of AC, or for a good deal of time "until cancelled", or a renewed time limited thing. That's esoteric and the marketplace can do whatever it wants. But every user would need to specify this in order to turn on their electricity, or else, they won't have any electricity to turn on.

    Said user could simply buy a box that has all of this programmed into it, or get it from a company that has said box in their possession (such as a big old utility). Or every motor, compressor, light bulb, etc., can be its own negotiator. That's esoteric, unimportant, irrelevant and doesn't matter EXCEPT THAT ALL OF IT SHOULD BE ALLOWED NO MATTER WHAT because the system will fall apart with artificial limitations. Each one just as much as it is behind a breaker is also behind another market participant in the grid chain, so it can't go wild as such.

    Some inverters only need software upgrade to do all that is necessary for this. Others would only be able to sell a lower quality (more sinusoidal, less "power-factored") waveform. The one change in all of this is that secure synchronous communications would need to be implemented systemwide across the grid. I'd say best is a fiber to every node, but for now, various stop-gap bandaid measures as radio through grid and piggybacking over communication company lines will work fine while the half-century buildout of fiber takes place (it can just be fiber on every wire until it reaches critical mass to connect a few last holdouts and that's done, and meanwhile being able to contract from other communication-only facilities will be easier and easier as this is more facilitated). It's not a lot of information: a bit of synchronous waveform timekeeping and marketplace negotiation with security. Sure, it's more than a 300bps modem, but really, it can be very lightweight.

    For experiments, such as in Hawaiian islands, it's OK to use radio transmissions to try this stuff out because after a programming experimentation and debugging period they can hardwire the communications. Long term, radio comm is bad since it is sabotage prone.

    Little grannie with her little solar panel collection and "that box in the garage" won't need to know these details, but could learn as much as she wanted about them. What matters is that she would be a master of frequency and kvar, of power factor and marketplace needs for various load waveforms. Regardless of her solar, wind, battery, hydro, etc., profile.

    And if by chance grannie gets weird and (a) doesn't maintain some faulty equipment or (b) gets downright saboteur, the grid operators can come disconnect grannie from the grid. All is well. (As if it isn't obvious by now, "grannie" can be anybody, your grannie, my grannie, you, me, a man, a woman, a small business with roof space, the little microdam microhydro facility in the back woods at the stream, the big ass wind farm up on the hill, the $2G 2GW modern gas generator plant, etc.)

    The marketplace really can solve a lot of ills, and this is one of them.

    Also, companies with equipment like capacitors connected in the correct places in the grid can offer these elements for a price, too. There's capital and maintenance to install such elements, and they have efficiency levels when connected. That's the utility. If they offer this method to clean up the waveform, then it might be bought in lieu of inverters inverting the wanted waveforms.

    The element of enforcement, of marketplace fairness, would be set by the grid operator: "thou shalt not use power in one power load factor way that is incongruent with how you purchase it". After that, purchase and sales would be set upon the deliverance of the power factor they actually use. This would all be entwined of course in the cost of electricity, etc.

    Most of this stuff has been intuitively obvious to me for many decades, and I assumed that we would be seeing dynamic experiments at universities showing how to program this stuff efficiently. But seeing the weird ass "my frequency my frequency" gollum-like chanting from that Hawaiian island utility man using a large amount of Solar City panels makes me realize this obvious solution hasn't been turned on. It's time to develop and install it, now. Do it! I'm speaking directly to you, Elon Musk, et al.

    I'm looking for references. I watched a video about it on those islands. I believe it might be Kauai Island Utility Cooperative (KIUC). But maybe the vid was scrubbed because of the opportunity to fix a simple thing became SO OBVIOUS and Big Oil doesn't want that? I'll search further.

    Anyway, this is the little fly in the ointment that I thought would already be solved by now, but obviously hasn't.

    To the layperson: the message is simply this: if big utilities are using excuses to say distributed solar panels aren't ready for the grid, then they are LYING: of course they aren't ready, because big utilities have been BLOCKING them from BECOMING READY, and even worse, the types of people who are inventive enough to fix this have been diverted from their paths, some intentionally by big oil, others by WHO KNOWS WHAT, maybe stupidity. But it will be done. The question is when, not if. Let's all be encouraging it to be done now, and to be done in a marketplace that isn't highly artificially constrained. This needs engineers, not policies or paperwork. Distributed generation is the new norm coming, so our fantastic grid providers need to get used to that fact and help us help them do it right.
  2. Ulmo

    Ulmo Member

    Jan 19, 2016
    San Jose, California
    #2 Ulmo, Jan 23, 2016
    Last edited: Jan 23, 2016
    Just saw where some of this "frequency" complaining comes from. In reading about German PhotoVoltaic marketplace, they noted that some solar PV installations are being designed to shut down due to "high frequency", since this is an indication of oversupply. This is because rotational generation is sped up when load is shed quickly. In places where sun can be blocked by clouds, when the clouds clear, this would have a similar effect. But all of this in my mind is caused by an old fashioned way of floating frequency off of other generation in a sort of mob consensus format. That's an archaic way of setting frequency from my experience.

    I grew up with Dave Mills from UDel, who helped maintain NTP, Network Time Protocol. This is an autocratic distributed master time method that seeks to find the One True Time, by using various heuristics to arrive at copying Colorado's government timepieces as accurately as possible (often with radio intermediaries, but today with exhaustive fiber internetworking could probably just as easily be done through glass). While probably not nearly as purpose-engineered as it would need to be for power frequency, it shares one thing in common with the modern way of doing things: finding a less ad-hoc way of figuring out what time (or in the case of electricity, what sine sync) you're in.

    I think once large distributed generation proportion fields in USA (by this I mean places now like Hawaiian islands where solar PV is having huge effects and soon the rest of us incrementally) realize that the generators set the frequency, and are the masters of the frequency, and that everybody who generates are The Generators (i.e., what we used to call The Utility*), they will realize they all have to sync to a common source, probably an out-of-band time-synchronous communication method with strong authentication for security measures. Probably the same source Dave Mills (and we all now) used. Yeah, maybe some geeks piggyback on GPS for a while to get it up and running and stuff, but you always need some redundancy (a local precise ticker pool, fiber connected, ratified as appropriate with deep sea cables, etc., not to mention severance policies if the governments split (e.g., Hawaii and Colorado start fighting each other)). In this environment, those solar PV farms in Germany that use frequency as a method to know about oversupply won't shut down because the frequency will be too stable even during oversupply events. Other methods would have to be used. (Or, do we want flywheels as canaries?)

    To USA grid-utility counterparts: however you used to manage frequency is how you need to manage it now, only, on a larger and more distributed scale. Mom and pop with their quiet little solar cells on their little roofs are now part of that. You have to sell them sine sync data. During startup, you might incentivize this by giving them the data and requiring it for interconnection.

    To CAISO: Don't make the mistake of thinking this is "just taken care of by the old way of doing things". You are the one holding the big wire. You can step up and say "And Henceforth The Upward Zero Axis Cross Point Of Phase 1 Shall Coincide With Atomic Time When The Second Strikes, With A Skew To Last Exactly 86,400 Seconds Of Equal Increments During Times of Second Elongation Or Shrinking By Atomic Time Bureaucrats (note: this is not leap seconds; leap seconds wouldn't hurt) (or does PG&E use downward? #2? Peak? You'd have to go see if they already have a standard.)". I bet PG&E already defined it as something, so whatever they already chose, just choose that. I'd be surprised if it was much different from the thing I made up right now. And, it's entirely bureaucratic at the Beginning, so you'd have some fun Writing the Decree. But, you'd have to find out what they're synced to before you mess it up by choosing a different one from the standard. But that's not the point, after all: anybody (smart enough) can select a starting point. The hard part and the point of everything I'm trying to say is that then you have to distribute and enforce this standard. There is a market idea for this in the post I responded to which is one of the best ideas, but a starting point is always going to be that everybody starts their phase as described. All those old utility experts that made a non-floating system here in our country can describe to you their methods, and something analogous to those ideas would have to make it into all the solar generators, wind generators, etc.

    * In this future (that we are headed to now), The Utility actually is just a grid and market broker that does paperwork with governments too. In the past, The Utility also took care of generation. In the future, they will only take care of a variable PORTION of generation, not The generation.
  3. Cyberax

    Cyberax Member

    Jul 28, 2015
    Seattle, WA
    Ugh. So much fail in this topic.

    And no, you can't just ignore frequency and power factor, even if your consumers can tolerate it. That's because transformers can not tolerate large swings of frequency.

    Even in a libertarian paradise where each lightbulb extorts the nearby electric outlets for power, utilities will still have to provide good power because otherwise a lot of important consumers will have to be dropped.
  4. Ulmo

    Ulmo Member

    Jan 19, 2016
    San Jose, California
    A more mature article was written about this over half a year ago, discussing the same issues using a higher level of language and a more company-centric point of view, yielding highly relevant reading:

    Californias Distributed Energy Challenge: Sharing the Data @ Greentech Media

    There are a number of relevant articles at links from here:

    Grid Resiliency -- Distributed Energy Resources @ SolarCity

    - - - Updated - - -

    Perhaps it wasn't clear to you that the idea-pulled-out-of-a-box proposal made above would work by each party offering items for sale in the market which would include the proper frequency and power factor, and it would be bid and agreed to on price (in rapid computer-speed fashion), thus the frequency and factor needed would be requested, found, bought, sold and provided for by a generator, the generator's inverter and dispatch matching the demanded item (frequency, power factor); the requested frequencies and pf's would be mediated by the grid services to be grid compatible, and thus, the bought and sold (provided) (generated inverted) electricity would converge to the right frequency and power factor, making the grid whole. Your response presumes the opposite of what was meant (which not so incidentally is a common tactic of nay-saying impedimentary lobbyists).
  5. FlasherZ

    FlasherZ Sig Model S + Sig Model X + Model 3 Resv

    Jun 21, 2012
    What happens when a bunch of IT guys try to talk about the frequency of the power grid? You get proposals to maintain a "master clock" like that used on the telecommunications network, with clock strata and "master sources" and such.

    Unfortunately, that's not what happens in the real power world. Frequency *must* be set by generator consensus because of mechanical turbines, unless you feel compelled to do very expensive double-conversion to stabilize it. You can use a master clock to govern the turbines, ensuring they spin at a target RPM, but that's a control plane vs. power plane separation and there is no need for some super-authenticated master stratum 1 clock. We're talking 60 Hz, not 1.5 Mhz, 125 MHz, 1.25 GHz.

    In an all-solar and all-wind world with battery storage, you could improve how the frequency is generated, but is it necessary? Meh.

Share This Page