Power Jack grid tie inverters

135

Comments

  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: Power Jack grid tie inverters
    To use my example, when ten SGPVs produce 40 amps of current on a branch designed for 15 amps max, I agree that an overcurrent condition exists. And when that happens, each fuse in each SGPV blows, removing the source of the overcurrent. It would be pointless for the grid breaker to trip since this would do nothing to address the source of the overcurrent--which is not the grid but the ten SGPVs. Are you saying that the fuses in the SGPVs will not blow? Or that they won't blow fast enough? Or that even if they blow, the overcurrent condition will remain?

    No, it doesn't work like that.
    Each SGPV sees only its own peak output, not the total current flowing through the wire. There's no way they can sense that in and of themselves.

    Essentially you'd be powering the one wire from several different sources. Each source (including the main breaker panel) protects the wire against exceeding the individual source peak current, but there is absolutely no circuit protection on the total current in the wire.
  • Sparkletron
    Sparkletron Solar Expert Posts: 71 ✭✭✭✭
    Re: Power Jack grid tie inverters
    Not really a matter of law so much as of physics.

    Oh there's quite a bit of law (politics) involved in electrical codes. For example, most homes do not have adequate grounding or protection from lightning, because such protection is immensely expensive. But we have something better than physics--we have actuarial tables. And so we perform the cost-to-benefit analysis and accept a certain amount of risk. But--again--what the USA considers an acceptable risk may differ from one locale to the next. Physics can give us the facts but it can't decide what an acceptable risk is.
  • Sparkletron
    Sparkletron Solar Expert Posts: 71 ✭✭✭✭
    Re: Power Jack grid tie inverters
    No, it doesn't work like that.
    Each SGPV sees only its own peak output, not the total current flowing through the wire. There's no way they can sense that in and of themselves.

    That's interesting, because these devices will not work unless they sense existing power on the line. That's one of the safety features. So you're saying that the SGPV can detect existing power on the line, but cannot know anything about the nature of this power? Neither the amps nor volts? How is this possible?
  • solar_dave
    solar_dave Solar Expert Posts: 2,397 ✭✭✭✭
    Re: Power Jack grid tie inverters
    Oh there's quite a bit of law (politics) involved in electrical codes. For example, most homes do not have adequate grounding or protection from lightning, because such protection is immensely expensive. But we have something better than physics--we have actuarial tables. And so we perform the cost-to-benefit analysis and accept a certain amount of risk. But--again--what the USA considers an acceptable risk may differ from one locale to the next. Physics can give us the facts but it can't decide what an acceptable risk is.


    And the physics are with 4 of these devices plugged in (16amps) and attached to a 15 amp house run mains breaker the possibility exists to have 31 amps of potential power on a 14 gauge wire which will burn if a large load is also placed on the line and no breakers or fuses will trip. It is as simple as that. Being non dedicated can cause a catastrophic wire failure. This is why there is NEC code for adding solar to your house. It is to protect you and your life & loved ones and your possessions.

    I am not saying these device can not be made safe if done properly, they probably can if one designs the backfeed system with proper code in mind. But you will play hell to get the utility and the local permitting agency to buy off on it.
  • solar_dave
    solar_dave Solar Expert Posts: 2,397 ✭✭✭✭
    Re: Power Jack grid tie inverters
    That's interesting, because these devices will not work unless they sense existing power on the line. That's one of the safety features. So you're saying that the SGPV can detect existing power on the line, but cannot know anything about the nature of this power? Neither the amps nor volts? How is this possible?



    Generally they detect volts and line frequency, but they can know nothing about the loads (amps) placed on the line. How could they? the loads do not run through the device.
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: Power Jack grid tie inverters
    That's interesting, because these devices will not work unless they sense existing power on the line. That's one of the safety features. So you're saying that the SGPV can detect existing power on the line, but cannot know anything about the nature of this power? Neither the amps nor volts? How is this possible?

    All grid tie systems shut down if they detect no AC power or power outside of acceptable parameters (too low, too high, wrong frequency). This is anti-islanding and prevents them from energizing the neighborhood in the event of an outage.

    To each of these units the over-loaded wire would appear as correct for Voltage, maximum current, and frequency. They would not shut down. The only part of the circuit that will "see" something is wrong is the over-loaded 14 AWG right before it fries.
  • Sparkletron
    Sparkletron Solar Expert Posts: 71 ✭✭✭✭
    Re: Power Jack grid tie inverters
    To each of these units the over-loaded wire would appear as correct for Voltage, maximum current, and frequency. They would not shut down.

    My understanding is that when a current is traveling along a wire and encounters a branch, the current divides evenly. If you have ten SGPVs all contributing current, each stream should divide upon encountering various branches--including the feed lines of all the other SGPVs. Since there are eleven branches (one for each SGVP plus the house branch), each SGPV's stream should encounter ten branches and contribute .4 amps to each branch. But since we have ten SGPVs all doing the same thing, then each branch ends up with 10 * .4 = 40 amps.

    Now if my understanding of what you're saying is correct, then I'm wrong, and each branch, upon encountering ten other branches (nine SGPV branches plus the house branch), doesn't divide it's current evenly, but instead contributes all of its current to the house branch and ignores all the other branches. Can you explain a bit more how this selective process works?
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: Power Jack grid tie inverters
    My understanding is that when a current is traveling along a wire and encounters a branch, the current divides evenly. If you have ten SGPVs all contributing current, each stream should divide upon encountering various branches--including the feed lines of all the other SGPVs. Since there are eleven branches (one for each SGVP plus the house branch), each SGPV's stream should encounter ten branches and contribute .4 amps to each branch. But since we have ten SGPVs all doing the same thing, then each branch ends up with 10 * .4 = 40 amps.

    Now if my understanding of what you're saying is correct, then I'm wrong, and each branch, upon encountering ten other branches (nine SGPV branches plus the house branch), doesn't divide it's current evenly, but instead contributes all of its current to the house branch and ignores all the other branches. Can you explain a bit more how this selective process works?

    Current only divides evenly if it branches to a parallel path. There is no parallel path here. There are several separate sources feeding into one wire. The current available from each source is limited to its maximum, but the current along the common wire can be the sum and total from all sources.

    It's like having a whole lot of on-ramps to the freeway: lots of cars coming in from all directions. Each ramp has light traffic on it, but when they all get on the common corridor you get grid lock: too much traffic trying to drive down too little road.

    Please don't take this the wrong way but your difficulty in understanding the problem is another problem with these plug-n-play devices: too many people assume that because it is sold and has a fuse in it it must be safe to use. so long as you don't do it wrong, it would be. But it's much too easy to do it wrong.

    BTW, most of the houses that have inadequate wiring in any form do so because they were built before the code required the grounding, et cetera. Any house built to code these days will have adequate wiring, including grounds. But "built to code" is the key; you can cheat on any building code aspect - and thus make the house unsafe.
  • Sparkletron
    Sparkletron Solar Expert Posts: 71 ✭✭✭✭
    Re: Power Jack grid tie inverters
    solar_dave wrote: »
    I am not saying these device can not be made safe if done properly, they probably can if one designs the backfeed system with proper code in mind. But you will play hell to get the utility and the local permitting agency to buy off on it.

    Exactly. And that's politics--not physics.

    I honestly don't know the exact nature of the overcurrent protection on these devices, but you moddies seem absolutely certain that they won't work if you combine enough of them. I don't have multiple SGVPs upon which to test any of this, so I'm just thinking in terms of batteries and what happens when you add additional batteries. That is something I could test and measure on my end.
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: Power Jack grid tie inverters
    Exactly. And that's politics--not physics.

    I honestly don't know the exact nature of the overcurrent protection on these devices, but you moddies seem absolutely certain that they won't work if you combine enough of them. I don't have multiple SGVPs upon which to test any of this, so I'm just thinking in terms of batteries and what happens when you add additional batteries. That is something I could test and measure on my end.

    The problem isn't with the inverters themselves, it's with plugging enough of them into the same wire so that wire has the potential to be overloaded. Each inverter can only sense its own current, not the collective total. If the wire only feeds the breaker at the box (as with a dedicated GT circuit) then an excess output will trip that breaker before the wire is overloaded. When you're able to plug in as many current sources as you wish that will feed unknown loads between the inverters and the breaker connecting to the main grid you've got the potential for a problem.
  • Sparkletron
    Sparkletron Solar Expert Posts: 71 ✭✭✭✭
    Re: Power Jack grid tie inverters
    Current only divides evenly if it branches to a parallel path. There is no parallel path here.

    So my house branch comes to a standard outlet. Let's say it has two sockets. You're telling me that each of those sockets is not connected to the house branch in parallel? Of course they are! And into each socket I plug in an SGVP. Now each SGVP is connected to the house branch in parallel. How could it be any other way? If the sockets of an outlet were connected in series, then neither socket would work unless they were both being used at the same time.
    It's like having a whole lot of on-ramps to the freeway: lots of cars coming in from all directions. Each ramp has light traffic on it, but when they all get on the common corridor you get grid lock: too much traffic trying to drive down too little road.

    I get the analogy. But electrons are not like cars eager to get on the freeway. They don't know which branch is the freeway. They encounter branches and divide--even if some of those branches are in fact "on-ramps" from other power sources.
    Please don't take this the wrong way but your difficulty in understanding the problem is another problem with these plug-n-play devices: too many people assume that because it is sold and has a fuse in it it must be safe to use. so long as you don't do it wrong, it would be. But it's much too easy to do it wrong.

    Believe me if I end up learning something it will make my day. And for what it's worth--I am listening and trying to give you moddies the benefit of the doubt.
  • Sparkletron
    Sparkletron Solar Expert Posts: 71 ✭✭✭✭
    Re: Power Jack grid tie inverters
    Each inverter can only sense its own current, not the collective total.

    This is really the only point of contention. I understand all about wire gauges, overcurrent, etc. The only part of this I don't grasp is how all of the feed lines are not experiencing the exact same amperage as all the rest (including the house branch). Somehow, the current from all the SGVPs avoid every branch but the house branch. How? Maybe I need to see a diagram...
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: Power Jack grid tie inverters
    So my house branch comes to a standard outlet. Let's say it has two sockets. You're telling me that each of those sockets is not connected to the house branch in parallel? Of course they are! And into each socket I plug in an SGVP. Now each SGVP is connected to the house branch in parallel. How could it be any other way? If the sockets of an outlet were connected in series, then neither socket would work unless they were both being used at the same time.

    Your separate inverters are not complete parallel circuits as with standard house wiring. They are individual circuits which interconnect at a common point; the potentially overloaded wiring.

    If you have multiple charge controllers hooked up to the same battery bank, each controller and its wires only sees the current it is producing. If the connections are made to bus bars or other common connectors which then go to the battery, the wiring from the common point to the batteries sees the total from all charge controllers connected. Three controllers at 20 Amps each equals 60 Amps total going to the battery. But from the controller to the common point (on either polarity side) is still only 20 Amps. Only from the common points to the battery (on either polarity side) do you see the full 60 Amps. If each controller has a fuse capable of handling the 20 Amps the wiring from controller to common point is safe, but there is no circuit protection between the common point and the batteries.

    With multiple plug-n-play inverters there is no circuit protection after the point where the sources come together. Thus it becomes possible to overload the wire after that point.
  • Sparkletron
    Sparkletron Solar Expert Posts: 71 ✭✭✭✭
    Re: Power Jack grid tie inverters
    Your separate inverters are not complete parallel circuits as with standard house wiring. They are individual circuits which interconnect at a common point.

    OK let me see if I can grasp this. The reason the current from one SGPV can't flow into the feed line of another is because there's no path to complete the circuit, because these inverters are not electrical loads.

    BTW, this means that charge controllers have the same issue, and all their overcurrent protections are for naught if not used properly.
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: Power Jack grid tie inverters
    OK let me see if I can grasp this. The reason the current from one SGPV can't flow into the feed line of another is because there's no path to complete the circuit, because these inverters are not electrical loads.

    BTW, this means that charge controllers have the same issue, and all their overcurrent protections are for naught if not used properly.

    Nope. Sorry. Still not right. The charge controller's circuit is through the battery and back the other side. But the only current it is "aware" of is what goes in and out of it specifically. Other controllers send their current through the common conductors (including battery) but not through the other controllers.

    Let's try some word pictures:

    Utility 15 Amp ==== OUTLET ==== OUTLET ==== OUTLET ==== OUTLET
    Normal household circuit, right?

    Utility 15 Amp ==== OUTLET ==== OUTLET ==== LOAD ==== OUTLET
    Here we've added a load.

    Utility 15 Amp ==== OUTLET ==== OUTLET ==== OUTLET ==== Inverters 10 Amp
    Here we have the inverters back-feeding the utility.

    Utility 15 Amp ==== OUTLET ==== OUTLET ==== LOAD ==== Inverters 10 Amp
    Here the load can draw from both power sources. If it is 22 Amps theoretically 11 Amps will come from each source.

    Utility 15 Amp ==== OUTLET ==== Inverters 10 Amp ==== OUTLET ==== LOAD 22 Amps
    This one is the problem. Current from the utility to the Inverters can not exceed 15 Amps, but from that point to the LOAD it can. The power coming from the utility does not go through the Inverters, so they don't "know" about it.
    Likewise the power coming from the inverters does not go through the Utility breaker. Thus from the point where the two power sources come together to where the LOAD is can be 22 Amps, but neither the circuit protection in the Utility box nor in the inverters sees this current.
  • solar_dave
    solar_dave Solar Expert Posts: 2,397 ✭✭✭✭
    Re: Power Jack grid tie inverters

    Coot hit the nail on the head. This is why a dedicated back feed is required by code, it controls the amps traversing the wire. Also the size of the main panel buss bar must be considered for a similar reason.

    Also an AC disconnect needs to be part of the circuit and usually a DC disconnect is also required before the inverters by code. These are there for fire department safety and be accessible from the outside of the house, clearly labeled.
  • ggunn
    ggunn Solar Expert Posts: 1,973 ✭✭✭
    Re: Power Jack grid tie inverters
    This is really the only point of contention. I understand all about wire gauges, overcurrent, etc. The only part of this I don't grasp is how all of the feed lines are not experiencing the exact same amperage as all the rest (including the house branch). Somehow, the current from all the SGVPs avoid every branch but the house branch. How? Maybe I need to see a diagram...
    The problem is not the branch current(s) but the current in the, er, trunk, where they merge. Each unit monitors its own current but none of them can see the current in the conductors where they all come together. Nothing is monitoring that current when there is a load elsewhere on that circuit, so it is very possible (easy, even) to have an unprotected overcurrent situation with insulation melting and fires starting. You say that you "understand all about wire gauges, overcurrent, etc." With all due respect, I don't think that you do. The reason why "all of the feed lines are not experiencing the exact same amperage as all the rest (including the house branch)", which you admitted that you do not grasp, is that the "house branch" carries the sum of all the inverter currents plus whatever load demand may be presented to the circuit in excess of those currents, and no single OCPD sees that total current.

    Code says that the total OCPD's feeding a conductor must not exceed 120% of the rated ampacity of that conductor. In this example, that is the sum of the breaker feeding the circuit from the service plus the sum of all the fuses or breakers in the renegade inverters plugged into that circuit. If the sum of all those exceeds 120% of the conductor's rating, then you are at risk of an overcurrent hot spot happening somewhere in that circuit. If it is a matter of law, then it's a good law, IMO. If it is an acceptable level of risk for you and you can get away with it, go for it, but if you start a fire your insurance company will (justifiably) deny the claim.

    EDIT: Here's a simple example. Say you have a 15A service breaker feeding a two receptacle circuit with correspondingly rated conductors. In the receptacle closest to the service, you have a plug-in inverter rated/protected at 15A. In the other receptacle, you plug in a load that draws 30A. The house breaker doesn't have a problem; it only sees 15A. Neither does the inverter OCPD; it only sees 15A. But the conductors between the first and second receptacle are carrying 30A. Normally (without the inverter's contribution), the house breaker would trip and protect the conductors, but in this scenario you may have just burned your house down.
  • Sparkletron
    Sparkletron Solar Expert Posts: 71 ✭✭✭✭
    Re: Power Jack grid tie inverters

    Those diagrams are very helpful.

    So even a single SGPV can create an uncontrollable overcurrent situation IF you locate it between the grid and the source (not at the end of a branch) and IF your load exceeds the max rated amps for the branch. And because of the difficulty in seeing through walls and knowing exactly which outlet is the last one on the branch, the average person is unlikely to implement the SGVP safely.

    Still, assuming that the load is a major appliance (it's drawing a ton of amps after all), it would almost certainly have its own overcurrent protection...

    GRID ===== OUTLET WITH INVERTER ===== OUTLET WITH LOAD

    Wouldn't this merely result in the load's fuse blowing?

    -S
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: Power Jack grid tie inverters

    Still, assuming that the load is a major appliance (it's drawing a ton of amps after all), it would almost certainly have its own overcurrent protection...

    In 50 years I've yet to see a major appliance with its own over-current protection.
    They put fuses in TV's, radios, stereos, computers ... all pretty light-weight stuff. But space heaters? Nah! What could go wrong there? :p

    One thing that has bugged me for years is the lack of circuit protection in extension cords. Easiest thing there is to overload, and they are one of the top causes of electrical fires.

    But even if the appliance did have protection, it would only be for its draw; not the total load on the household wiring. So again you have the same problem of being able to supply a section of wiring with more current than it is able to handle.
  • Sparkletron
    Sparkletron Solar Expert Posts: 71 ✭✭✭✭
    Re: Power Jack grid tie inverters

    Well certainly an appliance that needs 20 amps and expects to be on a 20-amp-rated branch will not have a problem when it detects 20 amps even if that represents an overcurrent for the branch. But if we're talking about following the rules, that 20 amp appliance is going to have a 20 amp NEMA plug, which is keyed and won't fit into a NEMA 15. Furthermore, the appliance's own overcurrent protection (even if it's merely a lowly fuse) will save the day in the event of a short-circuit.
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: Power Jack grid tie inverters
    Well certainly an appliance that needs 20 amps and expects to be on a 20-amp-rated branch will not have a problem when it detects 20 amps even if that represents an overcurrent for the branch. But if we're talking about following the rules, that 20 amp appliance is going to have a 20 amp NEMA plug, which is keyed and won't fit into a NEMA 15. Furthermore, the appliance's own overcurrent protection (even if it's merely a lowly fuse) will save the day in the event of a short-circuit.

    That only applies if a failure in the appliance itself is the cause of the overload.
    The problem we're looking at here is being able to put too much cumulative power source at one end of a wire and too much cumulative load at the other. Essentially "rewiring" the circuit so that it contains what are now inadequately sized conductors.
  • jeffkruse
    jeffkruse Solar Expert Posts: 205 ✭✭✭
    Re: Power Jack grid tie inverters
    Still, assuming that the load is a major appliance (it's drawing a ton of amps after all), it would almost certainly have its own overcurrent protection...

    GRID ===== OUTLET WITH INVERTER ===== OUTLET WITH LOAD

    Wouldn't this merely result in the load's fuse blowing?

    -S

    What if you plug in two loads, three loads, and decide to make toast? You generally don't know whats connected to the circut. Can you do it safely, sure, if you (and everyone in the house) understand what you are doing.+
  • Sparkletron
    Sparkletron Solar Expert Posts: 71 ✭✭✭✭
    Re: Power Jack grid tie inverters

    In my area, building code requires that all switched receptacles be installed upside down. It's very convenient. Perhaps something similar could be done for SGPVs. For example, if they were sold with a special keyed plug that won't fit any standard NEMA, and then an electrician is required to install the matching receptacle at the end of a branch...

    GRID ===== OUTLET WITH LOAD ===== OUTLET WITH INVERTER

    This would create the safe implementation that you were talking about. Of course it would raise costs (special plug, electrician), but it would still be considerably cheaper than what's involved with standard grid-tie.
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: Power Jack grid tie inverters

    Or in other words make sure the GT inverter is connected via a dedicated line, which is what the NEC says now.
    The whole point and problem with plug-in inverters is that it makes it just too easy to screw up and burn your house down.
    Not that there aren't a million ways for Joe I. Dunno to do that already. :p
  • Sparkletron
    Sparkletron Solar Expert Posts: 71 ✭✭✭✭
    Re: Power Jack grid tie inverters

    Well it wouldn't be a dedicated line because it would share the branch with the grid. The point is that these SGPVs are subkilowatt solutions. Too small for their own dedicated line. Too small for traditional grid-tie. Power companies would laugh if you wanted to grid-tie a 250W panel.
  • ggunn
    ggunn Solar Expert Posts: 1,973 ✭✭✭
    Re: Power Jack grid tie inverters
    Well it wouldn't be a dedicated line because it would share the branch with the grid.
    Which is precisely the problem. If your renegade inverter is running at full capacity, a load or combination of loads on the circuit it is plugged into has that capacity plus the full rating of the service breaker available before the breaker will trip. If that summed current exceeds the ampacity of the conductors, the service breaker can't protect them.
  • solar_dave
    solar_dave Solar Expert Posts: 2,397 ✭✭✭✭
    Re: Power Jack grid tie inverters
    ggunn wrote: »
    Which is precisely the problem. If your renegade inverter is running at full capacity, a load or combination of loads on the circuit has that capacity plus the full rating of the service breaker available before the breaker will trip. If that summed current exceeds the ampacity of the conductors, the service breaker can't protect them.

    That is why they will never get a UL rating.
  • solar_dave
    solar_dave Solar Expert Posts: 2,397 ✭✭✭✭
    Re: Power Jack grid tie inverters
    . Power companies would laugh if you wanted to grid-tie a 250W panel.

    Not really, they are done all the time with Enphase micro-inverters and a proper dedicated to code install. It is just poor economics to do just one, where the costs to permit and added the proper disconnect equipment are involved as they should be.
  • Sparkletron
    Sparkletron Solar Expert Posts: 71 ✭✭✭✭
    Re: Power Jack grid tie inverters

    I distinctly recall reading that some power companies require at least 1 kilowatt. But the fact is I just called my own utility and they confirmed that there is no limit with them. In any case, it's a lengthy and expensive process and I agree makes no sense in the subkilo range.

    I thought the point of micro-inverters was not to facilitate subkilo grid-tie, but to use many redundant inverters in lieu of one large inverter. That way if one panel goes offline, it doesn't bring down the entire system. Likewise, if one micro-inverter has trouble, the other inverters are still working. At the end of the day, the sum of all the micro-inverters is typically in kilowatts.
  • solar_dave
    solar_dave Solar Expert Posts: 2,397 ✭✭✭✭
    Re: Power Jack grid tie inverters
    I distinctly recall reading that some power companies require at least 1 kilowatt. But the fact is I just called my own utility and they confirmed that there is no limit with them. In any case, it's a lengthy and expensive process and I agree makes no sense in the subkilo range.

    I thought the point of micro-inverters was not to facilitate subkilo grid-tie, but to use many redundant inverters in lieu of one large inverter. That way if one panel goes offline, it doesn't bring down the entire system. Likewise, if one micro-inverter has trouble, the other inverters are still working. At the end of the day, the sum of all the micro-inverters is typically in kilowatts.

    Yes all of that is true plus installs where shading is a problem, just the point was that each panel has its own inverter but all the installation requirements exist and the overall system install does much better with scale.