Extending hours of operation through charging batteries

igor1960
igor1960 Solar Expert Posts: 85 ✭✭✭✭
I am looking for some advice from the experts on the design of the following GRID-TIE system.
The purpose of the system is to provide around 400wt per hour to cover minimum electric consumption by the house during both sun and dark hours and extend those “dark” hours for as long as possible, while not having NEM activated, thus providing minimum absolute immediate consumption reading on the main Meter.
So, I came up with the following design that consists of 2 separate banks of Solar Panels: Combiner1 and Combiner2.
Attachment not found.
During Sun hours (production and consumption): Combiner1 sends power directly to Power Inverter, while Combiner2 sends power to Battery Bank through Charge Controller.
During Dark hours (consumption): Combiner1 is out of power, while Battery Bank (that has been filled during Sun hours) sends power to Power Inverter.
Switching of current flow from Combiner1 (during Sun Hours) to Battery Bank (Dark hours) to Power Inverter is implemented by 2 Schottky Diodes (D1,D2) and is based on obvious voltage rise/drop from/to 40V/0V in positive line coming from Combiner1, while Battery Bank floats from/to 28V/20V.

Now, several questions to experienced people:
1. Lightning/Grounding question: do I need to connect Negative Bus (-) to separate ground, or I can just connect it directly to AC Ground (see (?) on the diagram);
2. There is no warranty that Power Inverter will not consume more than 15AMP of current during dark hours, as this is MPPT inverter and who knows how it will behave with the Battery Bank at the source, as its different IV curve. Also, Power Inverter may even decide not to consume enough power that I want it to produce. So, what is the way to guarantee around ~15AMP of current flowing through the Power Inverter during dark hours (when it’s working only on batteries)

Thank you

Comments

  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: Extending hours of operation through charging batteries

    If you want to run grid-tie with batteries use the proper hybrid type inverter. Do not try to run a standard GTI from batteries. Those diodes have no place in the design either.

    Go look at some wiring diagrams of existing system types and learn what you're dealing with. You're in real danger of starting a fire with your half-knowledge.
  • bill von novak
    bill von novak Solar Expert Posts: 891 ✭✭✭✭
    Re: Extending hours of operation through charging batteries

    What's the purpose of this? A system that supports your loads as much as possible without ever sending power back to the utility? This won't do that. It will always try to feed back power regardless of other loads.

    Are you planning to always consume an AC load of around 400 watts somehow?

    Is the power inverter shown a GT inverter? If so, which one? One of the Chinese "plug and play" inverters?

    Also note that a GT inverter may not "like" batteries as voltage sources. They may be too stiff a source, or may be too low a voltage for the inverter to be happy with. (Many GT inverters expect voltages close to max power point - thus you'd be sending it a voltage that is only 66% of what it expects.)
  • igor1960
    igor1960 Solar Expert Posts: 85 ✭✭✭✭
    Re: Extending hours of operation through charging batteries
    A system that supports your loads as much as possible without ever sending power back to the utility?
    Yes
    Are you planning to always consume an AC load of around 400 watts somehow?
    Yes... Constant pond load of around 400wt
    Is the power inverter shown a GT inverter? If so, which one? One of the Chinese "plug and play" inverters?
    Does it metter? Let it be "plug and play" inverter, but sitting on a dedicated line, with plug completely removed.


    So, no way to do this "safely"?
  • igor1960
    igor1960 Solar Expert Posts: 85 ✭✭✭✭
    Re: Extending hours of operation through charging batteries
    If you want to run grid-tie with batteries use the proper hybrid type inverter. Do not try to run a standard GTI from batteries. Those diodes have no place in the design either.
    Could you please elaborate on what is wrong with those diodes. Aren't they the same as normally used "blocking diodes"?
  • vtmaps
    vtmaps Solar Expert Posts: 3,741 ✭✭✭✭
    Re: Extending hours of operation through charging batteries

    Perhaps you should read through this thread:
    http://forum.solar-electric.com/showthread.php?16309
    Some of the issues you need to deal with are discussed there, in particular grid tie without supplying power to the grid.

    --vtMaps
    4 X 235watt Samsung, Midnite ePanel, Outback VFX3524 FM60 & mate, 4 Interstate L16, trimetric, Honda eu2000i
  • bill von novak
    bill von novak Solar Expert Posts: 891 ✭✭✭✭
    Re: Extending hours of operation through charging batteries
    igor1960 wrote: »
    So, no way to do this "safely"?

    LBX mode (also called HBX in Outback inverters) is probably what you want. From an old Trace manual:

    =================
    Low Battery Transfer (LBX) mode is an alternative way of operating “off the grid” using the utility for backup
    power instead of a generator. The system essentially operates as a stand-alone power system,
    independent of the utility grid. When the system is no longer able to keep up with the power requirements
    of the AC load, discharging the batteries to the LOW BATTERY TRANSFER VDC setting, the inverter
    connects to the utility grid. It then feeds utility power directly to the load and recharges the batteries.
    When the battery voltage reaches the LOW BATTERY CUT IN VDC setting, the inverter disconnects from
    the utility grid and once again operates the AC load from the batteries. Since power is never sold back to
    the utility, this configuration does not require utility approval.
    =======================
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: Extending hours of operation through charging batteries

    I've done several system designs for places where the grid is undependable and is basically used only as a 'generator'. That would be very similar to the intent here; rely on battery power to supply loads for as long as possible, make use of the grid power only when batteries are low. It is a simple adaptation of off-grid technology.

    I can't think of anyplace on Earth that has dependable grid power where it would be economically feasible to do this. In essence you would be using battery-based power most of the time, and that comes at a hefty cost.
  • igor1960
    igor1960 Solar Expert Posts: 85 ✭✭✭✭
    Re: Extending hours of operation through charging batteries
    I can't think of anyplace on Earth that has dependable grid power where it would be economically feasible to do this. In essence you would be using battery-based power most of the time, and that comes at a hefty cost.

    Why couldn't you imagine a situation, where for example small fish pond consuming constantly 400wt per hour for water circulation/aeration/UV sterilization has around 20 highly priced fishes, costing anywhere from $300 to $3000/per fish (koi & sturgent), where power interruption for more then 1hour would cost an owner far more then any investement in some kind of backup system?
    That is exactly what I'm talking about.
    So, in my original design I was contemplating on using grid-tie inverter, but in reality, having this design it is always feasible to add simple DC-AC inverter in parallel to Grid-Tie inverter, turning this inverter on during Dark Hours and send 400wt output directly to pond.
    I understand, that I might start with just a pond, as grid-off system. But idea was to eventually start small and then move slowly to become GRID Tied... at least have such possibility.

    With all due respect, but with all replies I'm getting that:
    1. I'm half an idiot having half-knowledge;
    2. Don't do this because of fire hazard;
    3. It's not safe;
    4. Those diodes have no place in that design (without any explanation);
    5. We are so "great", as we designed so many systems like this...
    6. That system is not economically feasible;
    7. Go get NEM (permit, application and etc. $1.5K), buy some expensive GTIs ($2K+), and only after that go and buy first panel (just for $100) -- and all that expence just for having pond backup for 400wt (possibly costing less then 2K overall)...


    But none of you really answered at least first question of mine:
    1. Do I need to connect Negative DC Bus with House Ground or not? I'm asking this, because I'm getting conflicting answers seraching internet about this.

    Thanx
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: Extending hours of operation through charging batteries

    Whether or not you connect the DC negative to Earth ground depends on a number of issues including exactly what inverter you use for whichever project you undertake and local regulations. For operation, no grounding is needed. To comply with the most recent NEC regs the DC side needs ground fault wiring, which does not connect negative to ground except indirectly.

    As far as the fish pond scenario goes, you would simply arrange for a UPS back-up for the pond pump. You do not need a GTI for this or even solar panels, merely an inverter capable of powering the pump and sufficient battery to see you through the expected outages with a method of recharging the batteries from grid when power is returned. There are a number of all-in-one UPS units such as that available at quite reasonable prices.

    In your original diagram the two diodes prevent A). the first array from feeding power to the battery and B). the battery from back-feeding the array at night but they do not prevent the inverter from drawing from the battery when there is power available from the array. Skipping the attempt to use a GTI as a battery-based inverter and using the that kind of inverter to begin with eliminates the wiring hassles and makes for a more functional system: inverter runs off battery, battery recharged or kept charged by solar array. Some off-gridders know the joy of being able to make use of opportunity loads when the batteries are fully charged and the sun is still shining; in essence this is what you are aiming for and is also how hybrid GTI's function; drawing power "directly" from panels to supply loads (and feed the grid in some cases).

    GTI's are not meant to power loads directly the way OGI's are; they simply add power to the wiring and it goes where it needs to go. Their output is not stable Voltage/varying current with demand like on off-grid unit (except for the hybrid type which may be considered an OGI).

    Yeah, I know; they're cheaper. But if you want a sure-fire formula for failure just try using any device for a purpose other than what it is intended for. You're trying to run before you can walk and that's likely to cause you a lot of disappointment and cost you a lot of money after which you still won't have something that works. How do I know? Because I have fixed oh-so-many systems that were a result of just such circumstances. If you don't want to take advantage of others' experience go ahead and do what you will, but do not make snide remarks about people who have that experience who try to short-cut you to results that will not waste your time and money.
  • igor1960
    igor1960 Solar Expert Posts: 85 ✭✭✭✭
    Re: Extending hours of operation through charging batteries
    In your original diagram the two diodes prevent A). the first array from feeding power to the battery and B). the battery from back-feeding the array at night but they do not prevent the inverter from drawing from the battery when there is power available from the array.

    Other then above, your points about safety, price and etc. are understood. Thank you.

    However, could you please explain above point that "two diodes...do not prevent the inverter from drawing from the battery when there is power available from the array".
    Maybe it's not clear, but each array output voltage during Sun Hours is around 40v, while Battery Bank voltage max. is around 28v. So, at D1 we have 40v and at D2 we have 28V. How Battery Bank could possibly back feed into Combiner1? Or Maybe you are talking about Combiner2?
  • bill von novak
    bill von novak Solar Expert Posts: 891 ✭✭✭✭
    Re: Extending hours of operation through charging batteries
    igor1960 wrote: »
    I understand, that I might start with just a pond, as grid-off system. But idea was to eventually start small and then move slowly to become GRID Tied... at least have such possibility.

    In that case consider the Outback GTFX line of inverters. You can start off in HBX mode (no grid tie) and then move to full grid tie just by changing the programming.
    1. I'm half an idiot having half-knowledge;
    2. Don't do this because of fire hazard;
    3. It's not safe;
    4. Those diodes have no place in that design (without any explanation);
    5. We are so "great", as we designed so many systems like this...
    6. That system is not economically feasible;
    7. Go get NEM (permit, application and etc. $1.5K), buy some expensive GTIs ($2K+), and only after that go and buy first panel (just for $100) -- and all that expence just for having pond backup for 400wt (possibly costing less then 2K overall)...

    It sounds like you asked a question and don't like the answers. That's a risk of asking a question - you might not get the answer you want to hear.

    If you want to build your circuit above, build it. You'll probably learn something about solar design. You probably won't accomplish your goal, which will mean overall you'll spend a lot more money. You'll make a lot of mistakes, which people here are trying to help you avoid. But if you want to make them - then go for it.
    1. Do I need to connect Negative DC Bus with House Ground or not? I'm asking this, because I'm getting conflicting answers seraching internet about this.

    Given the type of inverter you are using (cheap Chinese) it will depend entirely on the inverter. Many of these will self destruct if you ground the - line.
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: Extending hours of operation through charging batteries

    Array Voltage is not fixed, it is variable.

    Let's say you were to use a battery-based inverter here. You would then have the possibility of the array pushing Voltage at the inverter input up to Vmp which might exceed the max for the inverter (ask anyone with a cheap 12 Volt inverter who has tried to keep it running during an EQ cycle). Ideally with this dual-source you would want the nominal Voltages equal for consistent performance. If they are too far apart one or the other may fall outside of the inverter's operating range.

    PV's are a current source, batteries are a Voltage source. In a semi-isolated connection to the same unit the PV's will try to push current, but if the Voltage levels are similar the batteries are a far greater current source so the PV's will contribute little to the power demand; like putting an 'AAA' battery in parallel with a 'D' cell. Further the current from the array is dependent on the load placed on it, so the battery in taking priority for supply will reduce or eliminate the power from the array.

    Now consider skipping the semi-isolated set-up and supplying all array power through a controller. That in conjunction with the battery means a relatively stable input Voltage for the inverter. The current demand will vary with the load on the inverter, and so the panels will contribute current as much as they can because their Voltage potential is slightly higher than that of the battery (Vmp vs. resting state Voltage). This is normal operation for off-grid systems and also for hybrid systems where the grid is 'seen' as an unlimited current sink for the inverter.

    A GTI is meant to operate in a Voltage range. It already has an MPPT circuit on its input, and so it takes whatever power is available from the array and finds the maximum power point (V*I). It then adjusts its output to match detected line Voltage and frequency, and all power flows at that Voltage as a current into the 'unlimited load' of the grid. If you give it a battery for input the Voltage becomes stable and the current source unlimited potential until the battery discharges to a V below the inverter's cut-off. This could happen slowly or quickly, but there is no self-regulation of the current flow as there is with PV (Isc rating).

    Now, how would all this perform in the real world? A very good question to which the answer is that worst one of all "it depends". Depends on the exact specifications of all equipment used, that is; some of it may be more tolerant than others.

    The short form is that the end results can be achieved more easily, consistently, and cheaply with less danger just by selecting equipment meant to suit the need to begin with rather than trying to get something else to operate in a manner it was never intended to.
  • igor1960
    igor1960 Solar Expert Posts: 85 ✭✭✭✭
    Re: Extending hours of operation through charging batteries
    Array Voltage is not fixed, it is variable.

    Let's say you were to use a battery-based inverter here. You would then have the possibility of the array pushing Voltage at the inverter input up to Vmp which might exceed the max for the inverter (ask anyone with a cheap 12 Volt inverter who has tried to keep it running during an EQ cycle). Ideally with this dual-source you would want the nominal Voltages equal for consistent performance. If they are too far apart one or the other may fall outside of the inverter's operating range.

    PV's are a current source, batteries are a Voltage source. In a semi-isolated connection to the same unit the PV's will try to push current, but if the Voltage levels are similar the batteries are a far greater current source so the PV's will contribute little to the power demand; like putting an 'AAA' battery in parallel with a 'D' cell. Further the current from the array is dependent on the load placed on it, so the battery in taking priority for supply will reduce or eliminate the power from the array.

    Now consider skipping the semi-isolated set-up and supplying all array power through a controller. That in conjunction with the battery means a relatively stable input Voltage for the inverter. The current demand will vary with the load on the inverter, and so the panels will contribute current as much as they can because their Voltage potential is slightly higher than that of the battery (Vmp vs. resting state Voltage). This is normal operation for off-grid systems and also for hybrid systems where the grid is 'seen' as an unlimited current sink for the inverter.

    A GTI is meant to operate in a Voltage range. It already has an MPPT circuit on its input, and so it takes whatever power is available from the array and finds the maximum power point (V*I). It then adjusts its output to match detected line Voltage and frequency, and all power flows at that Voltage as a current into the 'unlimited load' of the grid. If you give it a battery for input the Voltage becomes stable and the current source unlimited potential until the battery discharges to a V below the inverter's cut-off. This could happen slowly or quickly, but there is no self-regulation of the current flow as there is with PV (Isc rating).

    Now, how would all this perform in the real world? A very good question to which the answer is that worst one of all "it depends". Depends on the exact specifications of all equipment used, that is; some of it may be more tolerant than others.

    The short form is that the end results can be achieved more easily, consistently, and cheaply with less danger just by selecting equipment meant to suit the need to begin with rather than trying to get something else to operate in a manner it was never intended to.

    Point well taken and agreed. I was thinking about the same possible voltage fluctuations due to MPPT algorithm. Though, mine GTI p&P inverter works in range of 22v-60v (not 10.8-30 as usual does). I've tested and connecting 28V DC battery through diode (D2) doesn't make any difference on output from it.

    Anyway, thank you for your input, as I understand and agree that involving the batteries is not the best way to go anyway... Was just thinking loud.