Another P&P Vendor - Can plug and play work in the US?

1356

Comments

  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    On my Phone right now. Just want to say the other folks here want to keep you, your family, and others reading SAFE.

    We have had these discussions many times before. To date, there has been no solutions. Including companies who told us nrtl listing is just around the corner or already signed off but is a trade secret. 1/2 a decade later, still waiting.

    And note that using quotes around "knowledgeable" experts in your initial posts sort of set an aggressive tone... Whether intentional or not.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • niel
    niel Solar Expert Posts: 10,300 ✭✭✭✭
    you love to put words into the mouths of others that was never said. i never said you can't do it as i said you need to hard wire it. now plugnplay implies simple connectivity as in standard plugs and outlets. you may be able to get an inspector to pass it if it used a non-standard connector and outlet inhibiting its use into a standard one. it needs inspected never the less whether your inspector is a moron or not and no matter what he charges for it unless you go guerrilla.
  • bill von novak
    bill von novak Solar Expert Posts: 891 ✭✭✭✭
    igor1960 wrote: »
    Imagine, that such inverters (using current plug) could be easily connected to dedicated circuit that has special CB. What's wrong with it?

    It could be easily misused in a dangerous way. It is akin to a two-male-end extension cord. Sure, in some cases it _could_ be used safely. But you wouldn't want such things regularly used by the public, because the death toll would be high.

    As an alternative, why not propose a dedicated connector? There are plenty of candidates available. Then contractors could follow simple rules for installation of both the home wiring (dedicated circuits only to that type of connector) and inverter manufacturers could use that connector on their inverters. That would seem to solve both problems.
  • bill von novak
    bill von novak Solar Expert Posts: 891 ✭✭✭✭
    igor1960 wrote: »
    Again, don't want to be rude, but it doesn't look you are comprehending what I'm saying. Repeating and imagine:
    1. We have specially designed circuit breaker ("CB for P&P inverters" line). They are special: they are marked 30AMP, but in reality they will trip at 15AMPs (so it's standard 15AMPS breaker inside);
    2. In addition to 15AMPS breaker that "CB for P&P inverters" has a small electronics (oscilator) that would add some known and standard frequency to the line (for example 8khz) , so we would have 60hz major AC frequency plus for example 8khz oscilator frequency on voltage (phase);
    3. Now, we connect P&P inverter using standard plug to this line. Inverters electronics using FFT checks for presence of 8khz in the input voltage and if such frequency exists continue working, otherwise just displays "Wrong Circuit"...
    4. Other consumers connected to that circuit, would continue to work, as they have no knowledge about 8khz frequency to check.

    What is wrong with this proposal?

    How do you ensure that 8khz signal doesn't get on any other lines? There is a galvanic connection through the load center busbars to every other circuit in the panel.

    And how do you prevent the scenario we discussed earlier? Say two 15 amp A/C loads, some additional loads (say 5 amps) and 20 amps worth of GT inverters near the panel. Total load is 15 amps with everything on; no breaker trip. Air conditioners are cycled off/turned off at night so no breaker trip then, either. Load along one section of wire - 35 amps.

    The above is a systemic problem with any such scheme since there is no control over how much GT solar and how much load you put on the line. It is made worse by some anti-islanding circuits. As the GT solar generation increases, and the voltage rises, the GT's may start tripping off-line as their voltages exceeds the threshold. Then that clever homeowner turns on his A/C, voltage drops, GT inverters come back on - and supply 35 amps through 30 amp rated wiring.

    (This is an unlikely scenario. But there are 140 million homes in the US - so if this has even a .001% chance of happening, it will happen 1400 times.)
  • igor1960
    igor1960 Solar Expert Posts: 85 ✭✭✭✭
    How do you ensure that 8khz signal doesn't get on any other lines? There is a galvanic connection through the load center busbars to every other circuit in the panel.

    Injected frequency would be seen only after CB not before, that might compicate things, but there must be way to isolate this.
    Small technical issue. Solution might be: implement this oscillation on down current instead then on voltage...

    And how do you prevent the scenario we discussed earlier? Say two 15 amp A/C loads, some additional loads (say 5 amps) and 20 amps worth of GT inverters near the panel. Total load is 15 amps with everything on; no breaker trip. Air conditioners are cycled off/turned off at night so no breaker trip then, either. Load along one section of wire - 35 amps.

    I'm confused here as my original message #20 states that we have CB twice lower then our circuit threashold. So, don't see any problems then with you example.
    Hypothetically and according to you we have 15AMP*2=30AM + 5 AMP = 35AMP => our circuit breaker should be 40AMP
    All wires, outlets and etc should be rated at 80AMPS (again hypothetically). Then we obviously could connect up to 40AMPS of inverters to this line.
    That's it -- that what I was saying...
  • solar_dave
    solar_dave Solar Expert Posts: 2,397 ✭✭✭✭
    No one has brought up the fact that none of the plug and play inverters available (with plug on end) are UL listed. Using them guerrilla will be a violation of your agreement with the utility for service and with the local AHJ for permitting. Connecting them in a method not approved by the NEC will also be a violation. These violations will invalidate your fire insurance policy so even if the device didn't cause a fire they may not pay off on the policy, they will look for any reason not to pay off, they are insurance companies in the business of making money.

    All the above reasons would keep a judicious person to NOT do these types of connections.

    If you have a dedicated 30 amp 240V circuit and a 200 amp service panel why not pull a permit and wire according to the NEC? A simple small AC combiner box with 2 15 amp breakers installed could be used with hard wired solar, assuming you have met the permitting requirements which will probably include panel grounding using #6 bare wire to the grounding source.

    This is not a place to cut corners on cost.
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    I found a couple old threads from two and three years ago (first thread here, second referenced in this thread too). Please do not adopt the tone there--It ended up locking that thread.

    Regarding using power line communications (8kHz or whatever)... Given that AC power mains can contain upwards of 50 kHz of "significant" power spectrum (mostly from non-power factor power supplies on AC mains--my guess). Filtering 8 kHz to limit "out of circuit" communications is probably not practical (without messing up the upper harmonics)--But going to a higher frequency is probably not an issue if you want to try it.

    It would be an interesting idea to put a "pilot tone/comm signal" on your "special circuit... Have a controller add the current from two (or more) power sources and modulate the GT Inverter output such that there is never more than 12 amps on the 15 amp circuit. Then you have to look at a "fail safe" (GT P&P shuts down if no tone, modulates GT output based on total current into the circuit--etc.). Will get a little complex (RMS calculations for non-linear current wave forms--Should not matter what the voltage or phase angles are--Wire/Breakers only "care" about total RMS current and not phase angles and such).

    And Micro GT inverters (Enphase at least--won't work here as they are 240 VAC devices) already use a power line communications network--Logging/control of micro inverters). Integrating with an already UL Listed GT inverter would be pretty interesting project (I think the present networking protocol is encrypted to some degree).

    Could it be done? Probably. Would it be "plug and play" by a consumer--No... Would need to pop open a main AC panel, identify the circuit, possibly move the breaker to the "bottom" of the AC mains bus. Figure out which brand/model, if ganged/multipole breakers, old aluminum wire installation, torquing breaker screws to correct torque... etc.

    Again, not easy. And not terribly expandable beyond 12 amps @ 120 VAC (US)... I just don't see a market if you have to involve an electrician, building permits, inspectors, Utility+Their rules, etc.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • bill von novak
    bill von novak Solar Expert Posts: 891 ✭✭✭✭
    igor1960 wrote: »
    Injected frequency would be seen only after CB not before, that might compicate things, but there must be way to isolate this.
    Sure, you could add a filter between the breaker and the busbar. Now you're adding cost, size and inefficiency though. (As with most engineering problems, you can solve any two of them.)
    Small technical issue. Solution might be: implement this oscillation on down current instead then on voltage...
    So modulate current? Tougher to do - and I don't think you'd want all your appliances singing an 8khz tune!
    I'm confused here as my original message #20 states that we have CB twice lower then our circuit threashold. So, don't see any problems then with you example.
    Hypothetically and according to you we have 15AMP*2=30AM + 5 AMP = 35AMP => our circuit breaker should be 40AMP
    Right. It should be - but the homeowner doesn't know that. He just plugs stuff in. So for our 15 amp (secretly 30 amp) circuit, he is once again overloading the wiring. Circuit breakers protect against stupid people as well as faults.

    And now you have a second problem. That 15 (secretly 30) amp circuit is now flowing through all your outlets rated for 15 amps. So some guy plugs in two hairdryers to the same outlet. (He uses a splitter so his wife and daughter can use them at the same time; he's so sick of the arguing and the shouting and the complaining . . . .) Load is now 28 amps flowing through a 15 amp outlet. That 15 amp breaker isn't going to blow because it's really a 30 amp. What happens to the outlet? If you're lucky it just gets hot. If not . . .
    All wires, outlets and etc should be rated at 80AMPS (again hypothetically). Then we obviously could connect up to 40AMPS of inverters to this line.
    Then he adds two 50 amp loads and 40 amps worth of inverters. Same problem. You can keep increasing wire gauges relative to breaker size until the odds of a problem are so low that UL is willing to work with you; maybe this 80 amp example works to guarantee you won't have a problem 99.99999% of the time. But is anyone going to pay for a system that uses 4 gauge wiring instead of 14ga Romex for the home runs?

    Here's another idea for you - smart outlets. Any run that is intended for this use has outlets/switches/devices that contain internal circuit breakers that trip if any _segment_ of the run sees more than the rated load. So a 15 amp breaker and 15 amp breakers at every load point (including any junctions.) This protects you from the issues above - but again either you replace every outlet and switch in the US with one of these or you go with a different geometry outlet so inverters can only be plugged into the new type. The nice thing here is now you can replace the plugs on your appliances and use them with the new outlets as well.
  • igor1960
    igor1960 Solar Expert Posts: 85 ✭✭✭✭
    Right. It should be - but the homeowner doesn't know that. He just plugs stuff in. So for our 15 amp (secretly 30 amp) circuit, he is once again overloading the wiring. Circuit breakers protect against stupid people as well as faults.
    .......
    .......and etc...

    It's completely opposite: even though we have 30amp circuit -- homeowner, doesn't know that. It is only known by electrician or whoever installed this. Homeowner instead knows that it's 15AMP circuit, as it is protected by 15AMP CB. So, everything you are saying is not valid, as homeowner will allow to connect max 15amp load...
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    I guess I do not understand...

    You have a 15 amp CB, 10 AWG cable, and some sort of socket.

    What is "special" about the AC socket? And how much current is available from the socket? Is there 15 amps (CB) plus GT inverter available current?

    Or is this based on your idea of 8kHz pilot frequency and a GT inverter that controls its output current based on the amount of current flowing through the "new intelligent CB"?

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • igor1960
    igor1960 Solar Expert Posts: 85 ✭✭✭✭
    Here's another idea for you - smart outlets. Any run that is intended for this use has outlets/switches/devices that contain internal circuit breakers that trip if any _segment_ of the run sees more than the rated load. So a 15 amp breaker and 15 amp breakers at every load point (including any junctions.) This protects you from the issues above - but again either you replace every outlet and switch in the US with one of these or you go with a different geometry outlet so inverters can only be plugged into the new type. The nice thing here is now you can replace the plugs on your appliances and use them with the new outlets as well.

    How about another and extremely simple idea:
    What if each P&P inverter has an algorithm that turns output off once in 1 second for 1 pulse of Vac. So, this would be done once in a second for 1 wave of AC frequency for example. Easily achievable with just 1/60 loss in power, which can be compensated on another remaining 59 cycles through use of small capacitor on DC side.
    1/60 of a second is around 0.017s=17ms <= exactly the time for CB to trip...
    Now if we have some kind of sync between multiple P&P inverters in the whole house: those that are sitting on the same circuit will do the same. That SYNC could be easily achieved by first P&P inverter introducing some high frequency pulse on the Vac wave. Next connected P&P inverter at startup will be looking for that pulse and if not found, that would indicate that this is the first one. Alternatively, some RF frequencies could be used for SYNC: like ZigBee from the meter and/or house wireless network phone/wi-fi and etc. Meaning that all P&P inverters will be in sync with each other.
    So, for one full cycle of Vac we would have only active LOAD passing current through the circuit breakers in the house => if ithe current is more then allowed CB will trip. Basically, same as today without P&P inverters installed, but just for 1/60th of a second...

    What would you say then? Would this work?

    So, if someone comes up with this implemenation and gets UL certification: would you guys be happy? Or you have some other objections?

    In addition to be UL listed there would be GFCI CB reuirement printed in documentation. Would this work?
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    igor1960 wrote: »
    How about another and extremely simple idea:
    What if each P&P inverter has an algorithm that turns output off once in 1 second for 1 pulse of Vac. Easily achievable.

    This may cause issues with PF correction requirements for equipment. I had problems meeting instanious power factor requirements for large communications systems that had variable power requirements (i.e., PF correction circuitry makes instantaneous correction in AC input based on DC power output requirements. Fortunately, the large computer system "averaged" out changes in DC power usage and therefore passed the qualifications tests.

    I would worry that dropping a 1/2 cycle will cause the breaker PFC qualifications to fail (not that CB were ever tested for PFC before).

    You will have other problems too... 1/2 cycle dropouts is just about the limit that a "typical" PC will carry through at low line before it power fails the DC side of the power system.

    Another issue is that for non-PFC switching power supplies, you will get higher surge currents through the rectifier/high voltage capacitor to make up the missing 1/2 cycle. Additional stresses are not something that you want.

    Missing 1/2 cycle will probably trip UPS systems to start converting (detect 1/2 cycle dropout and start conversion).

    You might cause buzzing in inductors/windings/sheet metal, and cause circulating currents in transformers (much like MSW inverters are known to do). And chopping the same same polarity 1/2 phase will case a DC offset current in the AC current--DC currents in AC transformers is not a good thing at all (DC Saturation currents are much lower than AC). So, you will have to have the circuit "toggle" the clipped phase so that DC average current is zero.
    So, this would be done once in a second for 1 wave of AC frequency for example. Easily achievable with just 1/60 loss in power, which can be compensated on another remaining 59 cycles through use of small capacitor on DC side.
    1/60 of a second is around 0.017s=17ms <= exactly the time for CB to trip...

    Standard Mains Transformers have current averaging. For example, the average surge current for an Induction Motor without any sort of "soft start"/Phase Correction is about 5x running current. I have also measured disk drives (Fujitsu Eagle 500 MB from the late 1970's--Weighed something like 100 lbs--those were the days)--Had a 45 amp starting surge (measured with a current clamp on a storage scope).

    So, an average Circuit Breaker will never "see/respond" to a 1/2 or full cycle of "over current). For example, a Square D QO 15 amp breaker (trip curves) will not reliably detect any current fault that is shorter than 1/60 second (60 Hz power). And the trip range at 0.02 seconds is around 15-55 times rated current (225 to 825 amps). This is not going to have the sensitivity/resolution (in this style breaker) to detect/trip on any 1/2 cycle 18 vs 15 amp over current.

    The only way this could work is, possibly, a microprocessor controlled Breaker--Much more complex (and probably expensive) that the standard thermal/magnetic trip type breaker. And the likely-hood of "false trips" is high (starting surge current, noise from universal/brushed motors, etc.).
    Now if we have some kind of sync between multiple P&P inverters in the whole house: those that are sitting on the same circuit will do the same.
    So, for one full cycle of Vac we would have only LOAD passing current through the circuit breaker => if it's more then allowed CB will trip.
    WHat would you say then? Would this work?

    Nope. Will not work with today's available hardware presently used in standard AC panels for homes and small businesses.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • bill von novak
    bill von novak Solar Expert Posts: 891 ✭✭✭✭
    igor1960 wrote: »
    It's completely opposite: eventhough we have 30amp circuit -- homeowner, doesn't know that. It is only known by electrician or whoever installed this. Homeowner instead knows that it's 15AMP circuit, as it is protected by 15AMP CB. So, everything you are saying is not valid, as homeowner will allow to connect max 15amp load...

    What will stop him from plugging two hair dryers into one outlet as I listed above?
  • igor1960
    igor1960 Solar Expert Posts: 85 ✭✭✭✭
    BB. wrote: »
    This may cause issues with PF correction requirements for equipment. I had problems meeting instanious power factor requirements for large communications systems that had variable power requirements (i.e., PF correction circuitry makes instantaneous correction in AC input based on DC power output requirements. Fortunately, the large computer system "averaged" out changes in DC power usage and therefore passed the qualifications tests.
    There would be no dropping of any cycles on ACTIVE LOAD. I was talking about turning off current FROM P&P inverter. So, your active load which is parallel to inverter at this moment (just for 1 cycle) will take full current from main. So, there will be no any voltage or current drop on ACTIVE POWER due to this interruption.
    BB. wrote: »
    So, an average Circuit Breaker will never "see/respond" to a 1/2 or full cycle of "over current). For example, a Square D QO 15 amp breaker (trip curves) will not reliably detect any current fault that is shorter than 1/60 second (60 Hz power). And the trip range at 0.02 seconds is around 15-55 times rated current (225 to 825 amps). This is not going to have the sensitivity/resolution (in this style breaker) to detect/trip on any 1/2 cycle 18 vs 15 amp over current.

    Agree here. But you are forgeting that we repeat that sequence each second. So, this is something to investigate: maybe it will trip after N seconds, which is OK if N is relatively small.
    BB. wrote: »
    Nope. Will not work with today's available hardware presently used in standard AC panels for homes and small businesses.

    I'm not talking about "presently used", I'm talking about how to make changes to P&P inverter it will become safe.
  • igor1960
    igor1960 Solar Expert Posts: 85 ✭✭✭✭
    What will stop him from plugging two hair dryers into one outlet as I listed above?

    Nothing. But that's not a point. The question is would the CB trip or not and if not would wires, receptacle withstand the load. And if they do it's OK.
  • igor1960
    igor1960 Solar Expert Posts: 85 ✭✭✭✭
    BB. wrote: »
    So, an average Circuit Breaker will never "see/respond" to a 1/2 or full cycle of "over current). For example, a Square D QO 15 amp breaker (trip curves) will not reliably detect any current fault that is shorter than 1/60 second (60 Hz power). And the trip range at 0.02 seconds is around 15-55 times rated current (225 to 825 amps). This is not going to have the sensitivity/resolution (in this style breaker) to detect/trip on any 1/2 cycle 18 vs 15 amp over current.

    BB you are so nice to me, and I agree with you on that one.
    To fix this problem I'm intoducing simplier algorithm:

    What if each P&P inverter has an algorithm that turns output off for example for 1 full minute, as soon as it sees unusual input Voltage drop. Because any ACTIVE LOAD turned on in parallel will cause some sudden voltage drop, which could be easily tracked by DSP of P&P inverter (accumulation of average and comparing with current value -- if difference is more then some threshold -> that is a signal that some parallel load was turned on and/or perhaps receiving unexpected jump out of perfect voltage sinusoid). If above condition found: P&P inverter stops current output for 1 minute for example.

    So, for 1 full minute we would have only active LOAD passing current through the circuit breaker on this line => if ithe current is more then allowed CB will trip. Basically, same as today without P&P inverters installed, but just for 1 minute...

    What would you say then? Would this work?

    So, if someone comes up with this implemenation and gets UL certification: would you guys be happy? Or you have some other objections?

    In addition to be UL listed there would be GFCI CB requirement printed in documentation. Would this work?

    How about writing Patent together?!

    In fact this power ON/OFF switch could be easily implemented as separate small device and we can just insert it into existing P&P inverter in between internal AC power supply and output receptacle...
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    igor1960 wrote: »
    There would be no dropping of any cycles on ACTIVE LOAD. I was talking about turning off current FROM P&P inverter. So, your active load which is parallel to inverter at this moment (just for 1 cycle) will take full current from main. So, there will be no any voltage or current drop on ACTIVE POWER due to this interruption.

    True--The main should provide the 1/2 cycle power and not trip a UPS. However, I do not believe you can have a GT inverter meet the PF>0.95 on the cycle per cycle requirement (as I remember from decades ago). (I was thinking of Hybrid Inverters where they have to do 1/2 power fail detection and switch to off grid mode--Does not apply here).
    Agree here. But you are forgetting that we repeat that sequence each second. So, this is something to investigate: maybe it will trip after N seconds, which is OK if N is relatively small.

    If you want the breaker to trip--It will trip somewhere round ~500 seconds minimum at 1.0 to ~1.25 times rated power. The high limit it >>10,000 seconds (if ever).
    I'm not talking about "presently used", I'm talking about how to make changes to P&P inverter it will become safe.

    The 1/2 cycle dropout would have to coordinated on all GT Inverters on the same branch circuit. DC average current could be an issue with short term (1/120 cycles) bias current (just don't know) to do the CB over current detection.

    I don't think it will be practical or meet the various code requirements...

    Without knowing the topology of the AC branch circuit (GT inverters should be at the opposite end of the run from the Main Breaker). But without tracing out wiring, that cannot be found out by "inspection" and if there are multiple sub branches (star wired), there is no "single end of the branch circuit" location. Gets us right back to how the NEC "uprated" Main AC panels--Main CB at top of bus and GT inverter feeds at bottom of bus to ensure that no section of bus has excessive current. And each tap from the bus has its own circuit breaker to limit fault current.

    I cannot see any "simple" method to allow a non-dedicated circuit (mixed GT Sources and Loads) could be used for Plug and Play GT inverters that addresses all the standard/abritrary permutations of standard home/small business wiring without active current measurement on a per wire connection (including J-boxes and outlets).

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    igor1960 wrote: »
    To fix this problem I'm introducing simpler algorithm:

    Keep it simple is always a good place to start. It does not take much to make to NEC safety confusing (grounding and neutral bonding is one of the common questions that generate long and complicated posts/threads).
    What if each P&P inverter has an algorithm that turns output off for example for 1 full minute, as soon as it sees unusual input Voltage drop.

    The only "voltage monitoring" by a "Circuit Breaker" so far (that I am aware of) in house panels is Arc Fault Detection--Basically the AFI listens for a spectrum that is frequently associated with arcs (hot to ground, hot across a poor connection, etc.). They subject to false trips (i.e., universal/brushed motor from a vacuum cleaner).

    I don't think you could reliably trigger anything (such as a GT 1 minute shutdown) on a suspicious voltage drop on an AC branch circuit...
    Because any ACTIVE LOAD turned on in parallel will cause some sudden voltage drop, which could be easily tracked by DSP of P&P inverter (accumulation of average and comparing with current value -- if difference is more then some threshold -> that is a signal that some parallel load was turned on). If above condition found: P&P inverter stops current output for 1 minute for example.

    People will have cows if their GT inverters shut down for 1 minute every time a water heater or A/C system fires up, or turns on their microwave (or their neighbor's A/C fires up).
    So, for 1 full minute we would have only active LOAD passing current through the circuit breaker on this line => if ithe current is more then allowed CB will trip. Basically, same as today without P&P inverters installed, but just for 1 minute...

    A 2x overload will trip the breaker in ~7-35 seconds--May not be "fast enough" to meet UL/NRTL/NEC current protection requirements for arbitrary wire heating (insulation, ROMEX, conduit, ambient temperatures, etc.).
    What would you say then? Would this work?

    Don't think you can do this based on voltage/frequency spectrum thresholds--Certainly not reliably in an arbitrary system.
    So, if someone comes up with this implementation and gets UL certification: would you guys be happy? Or you have some other objections?

    UL/NRTL Listing is certainly a start. You have to still review the installation to ensure that any particular installation meets the Listing requirements.
    In addition to be UL listed there would be GFCI CB requirement printed in documentation. Would this work?

    I don't know of any GFCI requirements for hard wired systems. For a plug and play system, yes, you would need a GFCI breaker (or GFCI outlet) because the AC panels have to be outside.

    The NEC is attempting to ensure (among other things), that every segment of wiring, when installed/configured per code, does not ever exceed its rated specifications--Through a combination of layered protective devices, ratings, and defined loads (and generation sources).

    That assumption was that there was one point of power source, and the rest were loads on a typical residential branch circuit. If there was a more complex requirement (such as a generator switch panel)--That was usually engineered by the mfg. and run through UL/NRTL inspection and tests for various possible fault conditions. If you have an "engineered" system, you can get away with some things that normal NEC would not allow--But that would be cost prohibitive to "re-engineer" a home to ensure that that specific branch circuit in that specific home was stable for mixed source/load operation.

    It is (almost?) always easier to install a new/dedicated circuit that meets the dedicated to GT Solar power than to trace down every twist and turn of the branch circuit just to install a GT inverter at the end of the string and install fused outlets at each point along the rest of the branch circuit.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • igor1960
    igor1960 Solar Expert Posts: 85 ✭✭✭✭
    ... check TI Inverter documenation here: http://www.ti.com/lit/an/sprabr5/sprabr5.pdf
    Second to last picture will show you 180 degrees shift between grid voltage and output current of the inverter at 500w production/load. Obviously, this confirms that output current of the inverter is 180 degrees shifted to circuit (main) downstream current (meaning they are working in opposite directions and therefore should be subtracted from each other on common loads and not added, as "professors" think)
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Technically speaking--Notice that channel 3 is actually "inverted" from its measured amplitude (see bottom of scope picture--INVERT "ON").

    Not that really matters, all is takes is the reversing the connections of the CT, M-FET/Hall current probe, or shunt to reverse the current polarity on the display (or use the "INVERT Trace" button).

    It just depends on how you define "positive and negative" current flow direction.

    In any case, if the 500 Watt inverter is outputting (for example) the positive sine wave, then the current flow has to be out from the AC inverter into an arbitrary load. If the current flow is into the Inverter as the wave from goes positive, then the Inverter is a Load--Something that would not be a GT inverter that we would want to use.

    It appears, for what ever reason, they chose to inverter the current direction to indicate the Inverter is outputting energy into the branch circuit rather than consuming it. The sign of the current flow is all "relative" to how you assume the device operates.

    If you look at page 20 section 4.2, you see the Hall transistor feedback loop--And their they have the Voltage in Phase with the Current wave form (and scope INVERT Trace "OFF").

    This does not do a good job of explaining power flow in an AC circuit and how phasing/polarity of current (0/180 degrees) "works".

    -Bill

    Note for exactness--Channel 3 is the Grid Voltage Channel that is inverted in the last set of pictures. Again--Whether the Voltage (ch3) trace or the Current (ch2) trace is inverted does not really matter... There was a sign (or effectively a phase) swap.
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • bill von novak
    bill von novak Solar Expert Posts: 891 ✭✭✭✭
    igor1960 wrote: »
    What if each P&P inverter has an algorithm that turns output off once in 1 second for 1 pulse of Vac. So, this would be done once in a second for 1 wave of AC frequency for example. Easily achievable with just 1/60 loss in power, which can be compensated on another remaining 59 cycles through use of small capacitor on DC side.

    Your fans, compressors and pumps would all "tick" once a second as they lost power for one cycle. Would probably be pretty annoying. And your stereos without extra filtering would "pop" once a second.

    But again this is beside the point. With current outlets this presents a problem for load and wiring protection.
    Now if we have some kind of sync between multiple P&P inverters in the whole house: those that are sitting on the same circuit will do the same. That SYNC could be easily achieved by first P&P inverter introducing some high frequency pulse on the Vac wave. Next connected P&P inverter at startup will be looking for that pulse and if not found, that would indicate that this is the first one. Alternatively, some RF frequencies could be used for SYNC: like ZigBee from the meter and/or house wireless network phone/wi-fi and etc. Meaning that all P&P inverters will be in sync with each other.
    So, for one full cycle of Vac we would have only active LOAD passing current through the circuit breakers in the house

    So for one full cycle you basically disconnect the breaker, during which time power flows from inverters directly to loads. Again, how does this help? If you have a case where you have 40 amps of load and 40 amps of solar on a 30 amp line, wouldn't the system you describe be perfectly happy even though it is overloading the wiring?
  • bill von novak
    bill von novak Solar Expert Posts: 891 ✭✭✭✭
    igor1960 wrote: »
    Nothing. But that's not a point. The question is would the CB trip or not and if not would wires, receptacle withstand the load. And if they do it's OK.

    That is indeed the question! And in that case here's what you see:

    30 amp breaker (labeled "15")
    14 amp x2 loads plugged into one outlet

    Load on one outlet - 28 amps
    Rating for one outlet - 15 amps
    Result - outlet overload
  • igor1960
    igor1960 Solar Expert Posts: 85 ✭✭✭✭
    30 amp breaker (labeled "15")
    14 amp x2 loads plugged into one outlet

    Load on one outlet - 28 amps
    Rating for one outlet - 15 amps
    Result - outlet overload

    Bill, you are constantly repeating not what in my proposal. Just concentrate, forget what you think you undertand (or imagine) and comprehend here with me:
    Again, In my proposal we have:

    Circuit line designed for 30AMP load: let's say #10AWG and all outlets for max. 30AMP, but instead of a Circuit Breaker that you usually put at 30AMP on that line, we install CB with just max.15AMP.
    Means, that whenever we have 15AMP current going through CB -- it will trip.
    So, we do not have "30 amp breaker (labeled "15")", as you think, but we really have 15AMP breaker.
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    igor1960 wrote: »
    Circuit line designed for 30AMP load: let's say #10AWG and all outlets for max. 30AMP, but instead of a Circuit Breaker that you usually put at 30AMP on that line, we install CB with just max.15AMP.
    Means, that whenever we have 15AMP current going through CB -- it will trip.
    So, we do not have "30 amp breaker (labeled "15")", as you think, but we really have 15AMP breaker.

    As I have already explained, so long as any wiring shared between the two power sources and the load is capable of taking the combined current potential it will work.

    You have gone 'round-about from wishing to use existing plug-n-play inverters to proposing a complete and largely impractical redesign of how they work. You have ignored explanations from people who do understand this whilst insulting them and implying you know more. As I said before, if you understand everything so well why do you ask the questions?

    Your attitude towards the other posters on this forum is not acceptable.
  • igor1960
    igor1960 Solar Expert Posts: 85 ✭✭✭✭
    BB. wrote: »
    Technically speaking--Notice that channel 3 is actually "inverted" from its measured amplitude (see bottom of scope picture--INVERT "ON").

    Not that really matters, all is takes is the reversing the connections of the CT, M-FET/Hall current probe, or shunt to reverse the current polarity on the display (or use the "INVERT Trace" button).

    It just depends on how you define "positive and negative" current flow direction.

    BB, you are the smartest here! I am really amazed.
    I was showing this to other "professors", who think that LOAD current and INVERTERS current should be added on common load lines, instead of subtracted.
    Seriously, what do you think?

    Assuming we are doing connection permitted by electrical company with central UL listed inverter. Everything sits on a separate 30AMP circuit. Perfect, and everything kosher.
    Assuming, this is all connected to 200AMP main panel with 200AMP main circuit breaker and 200AMP Meter.
    Now, we have the same problem here, as according to those "professors" eventhough we have 200AMP main CB, we might have 230AMP current somewhere inside our main panel, Right? According to them Yes: 200AMP load + 30AMP solar...
    Question now: did they upgrade their main panel? Is this required. I don't know, maybe, but this is laughable!!!
  • bill von novak
    bill von novak Solar Expert Posts: 891 ✭✭✭✭
    igor1960 wrote: »
    Circuit line designed for 30AMP load: let's say #10AWG and all outlets for max. 30AMP, but instead of a Circuit Breaker that you usually put at 30AMP on that line, we install CB with just max.15AMP.
    Means, that whenever we have 15AMP current going through CB -- it will trip.
    So, we do not have "30 amp breaker (labeled "15")", as you think, but we really have 15AMP breaker.

    OK. So here is my scenario:

    Grid tie inverters generating 15 amps total
    Two dryers plugged into same outlet drawing 14 amps each (28 amps total load)
    Rating on all outlets is 15 amps (i.e. standard NEMA 5-15R's)

    My questions for you:
    1) Will the breaker trip?
    2) If not, what will be the current drawn through the outlet that has the hair dryers plugged in?
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Igor,

    For the moment, forget AC--Think of a DC system. You have a 15 amp CB to a battery bank, and you have 15 amp alternator. Circuit:

    Battery->15 Amp CB->10 awg cable "A"-> "X" socket for 15 amp Alternator->10 awg cable segment "B"-> "Z" socket->appliance pigtail designed for 15 amp @ 120 VAC outlet

    I put a variable load on Z socket. Fire up the alternator and output 15 amps.

    The current flowing through 15 amp CB will be equal to [Alternator Output - Load at "Z" Socket]. Correct?

    Zero amp load at Z socket will give me 15 amps at CB. Correct?

    15 amp load at Z socket will give me "zero current" at CB and 15 amps in "B" cable and "Z" socket? Correct?

    30 amp load at Z socket will give me 15 amps at CB, 30 amps at Segment "B", and 30 amps at Z socket and appliance pigtail. Correct?

    More than 30 amp load at Z socket... 15 amp Alternator is already current limited, and 15 amp CB limit will be exceeded. CB will trip and alternator will shut down because of "anti-islanding" circuitry in alternator (not true in real life, but is "true" for GT based inverters). Correct?

    If this is a "standard" 120 VDC @ 15 amp branch circuit appliance, then its pigtail and UL Listing will be voided in this installation because of the 30 amp available power. If appliance is designed for a 30 amp circuit, then all is OK (other than it will not work if it requires more than 15 amps at night/less than full sun).

    Yes, your wiring and outlets are designed for "30 amps"--And nothing inside this DC House wiring example will exceed the 30 amp rating (ignoring the NEC 0.80 or 1.25x derating of wiring, outlets, and breakers).

    Do you agree with this as a DC example of your proposal? (forgetting the AC discussion for the above example).

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • solar_dave
    solar_dave Solar Expert Posts: 2,397 ✭✭✭✭
    igor1960 wrote: »
    BB, you are the smartest here! I am really amazed.
    I was showing this to other "professors", who think that LOAD current and INVERTERS current should be added on common load lines, instead of subtracted.
    Seriously, what do you think?

    Assuming we are doing connection permitted by electrical company with central UL listed inverter. Everything sits on a separate 30AMP circuit. Perfect, and everything kosher.
    Assuming, this is all connected to 200AMP main panel with 200AMP main circuit breaker and 200AMP Meter.
    Now, we have the same problem here, as according to those "professors" eventhough we have 200AMP main CB, we might have 230AMP current somewhere inside our main panel, Right? According to them Yes: 200AMP load + 30AMP solar...
    Question now: did they upgrade their main panel? Is this required. I don't know, maybe, but this is laughable!!!

    SO if you would take the time to read the NEC you would know that it is acceptable to have a 200 amp buss bar with up to a 40 amp solar backfeed. It is called the 120% rule, the breaker needs to be place at the opposite end of the buss bar. Now what do you do if you want to back feed more? You derate the main breaker to something less like 175 amps. That would allow up to a 65 amp backfeed (more probably an even breaker amount of 60 amps).

    BTW it is not laughable, it is the electrical code.
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    There is NO current without a load. If the inverter is back-feeding the grid then the grid is a load. If the inverter is powering a device and back-feeding then both are loads to the inverter. If the device draws more current than the inverter can supply the difference is made up by the grid. The power sources will always act in concert. That is how it is. Physics.
  • igor1960
    igor1960 Solar Expert Posts: 85 ✭✭✭✭
    You have gone 'round-about from wishing to use existing plug-n-play inverters to proposing a complete and largely impractical redesign of how they work. You have ignored explanations from people who do understand this whilst insulting them and implying you know more. As I said before, if you understand everything so well why do you ask the questions?

    I though it's not just FAQ board, but instead a usefull colaborative forum...
    Your attitude towards the other posters on this forum is not acceptable.

    God knows I didn't start that: Check one of your previous messages:
    Unfortunately the OP seems to have difficulty grasping the abstract examples used. Maybe if he tooka few years out to get an electrical engineering degree?

    Should I be sorry?