Opportunity Load

igor1960
igor1960 Solar Expert Posts: 85 ✭✭✭✭
Hypothetical situation:
Small PV system with a small inverter. SmartMeter in the house. NEM is not activated and will not be, as the purpose is just to drop electric consumption to minimal level and not sell it to electric company. Installed SmartMeter obviously integrates both received/consumed energy as both positive. No batteries, as this is complicated/costly.
During sun hours immediatet production of the system lets say 1kwt, while house usage at this moment is around 350wt. Both vary in time. However, SmartMeters ZigBee connected attachment shows immediate NET of both load/receive at -650wt (minus 650wt). Idea, is to divert this extra power to a small water heater and use the heater on AC side as an "opportunity load". With such "opportunity load" SmartMeter will show absolute minimum immediate load and/or ideally 0, so installed SmartMeter will show minimum integrated power consumption, as well as Electric company will never know that you have power "injector" (at least while you are NET positive).

Now question -- might be stupid:
Assuming small PC runs the whole setup like this and is able to receive NET load/receive on the SmartMeter:
Obvious questions:
1. What kind of water heater is needed? As I undertsand standard electrical water heater might not work, as it works by constantly turning on/off, thus either having max./min. load -- above setup needs heater that have heating elements capable to work with different available currents;
2. What kind of current/power regulating device should be used on the side of that heater? As I understand, such PC should be connected to some current/power limiting controller that would be dynamically regulated by computer in terms of how much power is allowed to go through it, so inline connected heater will consume that ammount of power and not more (something like a dimmer). So, assuming above example, computer sends a signal to allow 650watts of power and that device responds. Simple solution might be this dimmer http://www.smarthome.com/2477DH/SwitchLinc-INSTEON-Remote-Control-Dimmer-Dual-Band-High-Wattage-White/p.aspx -- 1kwt...

Any ideas. Obviously looking for cheap commercially available solution/parts.
Or, maybe there is better way to implement above, possibly installing/diverting current on a DC side?

Thank you

Comments

  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Re: Opportunity Load

    Here is one thread where this was discussed with some possible solutions:

    Grid tie inverter that does NOT feed into the grid


    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • mike95490
    mike95490 Solar Expert Posts: 9,583 ✭✭✭✭✭
    Re: Opportunity Load

    You will find (I suspect) that the cost of small solar with no backfeed, will only help you slightly. Most household loads (except HVAC and electric cooking) average less than 1,000 watts at any time. You only have a 5, possibly 6, hour window to zero your loads, and the rest of the time, you are still sucking on the Grid. You could pre-chill (or heat) your house, have a manual defrost freezer on a timer, to run only when sun is up, and a couple other loads. Unless you have a home office, and can shuffle clothes in and out of the washer and dryer, you likely don't have enough usable loads.
    And the Smart Meter will rat you out anyway, even though you don't feed the Grid, you are connected, and generally required to get permits and permission to connect electrical generating gear to it.
    Powerfab top of pole PV mount | Listeroid 6/1 w/st5 gen head | XW6048 inverter/chgr | Iota 48V/15A charger | Morningstar 60A MPPT | 48V, 800A NiFe Battery (in series)| 15, Evergreen 205w "12V" PV array on pole | Midnight ePanel | Grundfos 10 SO5-9 with 3 wire Franklin Electric motor (1/2hp 240V 1ph ) on a timer for 3 hr noontime run - Runs off PV ||
    || Midnight Classic 200 | 10, Evergreen 200w in a 160VOC array ||
    || VEC1093 12V Charger | Maha C401 aa/aaa Charger | SureSine | Sunsaver MPPT 15A

    solar: http://tinyurl.com/LMR-Solar
    gen: http://tinyurl.com/LMR-Lister ,

  • igor1960
    igor1960 Solar Expert Posts: 85 ✭✭✭✭
    Re: Opportunity Load
    mike95490 wrote: »
    You will find (I suspect) that the cost of small solar with no backfeed, will only help you slightly. Most household loads (except HVAC and electric cooking) average less than 1,000 watts at any time. You only have a 5, possibly 6, hour window to zero your loads, and the rest of the time, you are still sucking on the Grid. You could pre-chill (or heat) your house, have a manual defrost freezer on a timer, to run only when sun is up, and a couple other loads. Unless you have a home office, and can shuffle clothes in and out of the washer and dryer, you likely don't have enough usable loads.
    And the Smart Meter will rat you out anyway, even though you don't feed the Grid, you are connected, and generally required to get permits and permission to connect electrical generating gear to it.

    If the purpose is just to "fool" SmartMeter (so it shows abs min, but still positive immediate consumption)/Electric Company, then I do think it's not that difficult/expensive by just regulating/limiting current on Solar DC side.
    We can install non contact inductive AC current sensor in the main panel like this one: http://www.ebay.com/itm/Non-invasive-AC-current-sensor-SCT-019-200A-max-/150853751480, inline with AMP 0.33v->5v, and then the MOSFET like this one: https://www.sparkfun.com/products/10213, regulating solar DC current. We could probably be able to find other types of MOSFET(s), so we don't even need AMP (and therefore AMP and external power supply).
    Also, instead of current sensor we can possibly use voltage drop between Neutral and Ground bars in the main panel as a signal.
    Thus we would be able to regulate current on solar DC side based on load in the house.

    That shouldn't be expensive (probably less then $40), though we will not be utilizing Opportunity Load.
  • bill von novak
    bill von novak Solar Expert Posts: 891 ✭✭✭✭
    Re: Opportunity Load
    igor1960 wrote: »
    and then the MOSFET like this one: https://www.sparkfun.com/products/10213, regulating solar DC current.

    You do not want to try to drop that much power across a TO-220! He's talking about a 1000 watt source. With excellent heatsinking a TO-220 can dissipate about 75 watts. (And that's assuming that the GT inverter will work at all with a linear regulator changing its V/I characteristics.)
    Also, instead of current sensor we can possibly use voltage drop between Neutral and Ground bars in the main panel as a signal.
    You could, but that is very hard to calibrate, and inductive sensors are both cheap and accurate.
  • igor1960
    igor1960 Solar Expert Posts: 85 ✭✭✭✭
    Re: Opportunity Load
    You do not want to try to drop that much power across a TO-220! He's talking about a 1000 watt source. With excellent heatsinking a TO-220 can dissipate about 75 watts. (And that's assuming that the GT inverter will work at all with a linear regulator changing its V/I characteristics.)


    So, Bill: are you saying it is difficult? This would not work? Confused... What alterantive do you have in mind? Put 10 of them in parallel/series? Just trying to learn.
    BTW: Do you know if SCT-019 (http://www.ebay.com/itm/Non-invasive-AC-current-sensor-SCT-019-200A-max-/150853751480) changes its DC current direction (signal), based on AC current direction? And/or it just returns the absolute level of the AC current? I think this is important, Right?
  • bill von novak
    bill von novak Solar Expert Posts: 891 ✭✭✭✭
    Re: Opportunity Load
    igor1960 wrote: »
    So, Bill: are you saying it is difficult? This would not work? Confused... What alterantive do you have in mind?

    For the method to "tap off" some power - a buck converter driving a 220 volt element. Works as long as the Vmp of the array is greater than about 220 volts.

    For sensing, an accurate bidirectional power sensor (which involves a current sensor, voltage sensor, multiplier and RMS-DC converter) at the service entrance.

    A simplistic control scheme would close the loop around the power sensor, and drive the buck converter to prevent power from ever going negative. But this ignores the MPPT behavior of the grid tie inverter, which can interact with such a system and render it unstable. You would need access to the inverter control loop to ensure this does not happen.
    BTW: Do you know if SCT-019 (http://www.ebay.com/itm/Non-invasive-AC-current-sensor-SCT-019-200A-max-/150853751480) changes its DC current direction (signal), based on AC current direction? And/or it just returns the absolute level of the AC current? I think this is important, Right?

    Yes, it senses current in both directions. However, that in and of itself is insufficient to determine the direction of power transfer.
  • igor1960
    igor1960 Solar Expert Posts: 85 ✭✭✭✭
    Re: Opportunity Load
    A simplistic control scheme would close the loop around the power sensor, and drive the buck converter to prevent power from ever going negative. But this ignores the MPPT behavior of the grid tie inverter, which can interact with such a system and render it unstable. You would need access to the inverter control loop to ensure this does not happen.
    Any ideas on installing element on AC side (not to interfere with MPPT)?


    Yes, it senses current in both directions. However, that in and of itself is insufficient to determine the direction of power transfer.
    So, it doesn't switch output polarity on change in AC current direction, Right? Just shows absolute value. Not good. Do you know of any AC current sensor that does switch output polarity?
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Re: Opportunity Load
    igor1960 wrote: »
    Any ideas on installing element on AC side (not to interfere with MPPT)?

    In theory, turning on/off heating elements on AC circuits should not matter. In practice--It can. You may not notice the lights going dim for a faction of second and brightening 5 minutes later when the element turns off. However, cycle the element at 5-10 Hz--A very good chance that everyone will notice lights "blinking" around the home.
    So, it doesn't switch output polarity on change in AC current direction, Right? Just shows absolute value. Not good. Do you know of any AC current sensor that does switch output polarity?

    The polarity of the signal is the Phase of the signal in AC circuits. I think what you linked to (as I remember) was a current transformer--some of which had an internal precision resistor (a CT with open wires will generate 100's to 1,000's of volts and create a shock hazard and ruin the CT because of arc through in the insulation. CT with an integrated resistor, at least, stops the high voltage output (the damage from placing a CT on an active AC circuit with current flow).

    You just use a Analog to Digital converter to get a data stream (just like the voltage side measurement) that you can now do the math on (and there cannot be any delay between the two data streams--that will cause a phase error).

    In an ideal world, with PF=1, zero phase angle between Voltage and Current, you can set up the polarity for the current to always be positive with the voltage (current tracks voltage). (note, polarity/180 phase shift by just reversing the direction of the wire through the CT or flipping the CT output leads).

    In practice, you only have 0.95 PF for typical "good quality" AC power from a GT inverter--So, ~5% of the time the current flow will be "backwards" relative to the voltage (i.e., complex impedance verses simple resistance).

    This stuff is not trivial, but is doable (sometimes library functions may be available if you know were to look--There are one or two threads here by people who are doing home automation down to this level of detail).

    You can also do simple stuff (i.e., peak detection)... If the current peak is less than +/- 90 degrees wrt the voltage peak, you can estimate the direction of the power flow (I think--It has been several decades since I did any math at this level).

    If you don't want to do CT pickups, you can do similar stuff with Hall Effect sensors (on AC systems, I don't think you have the calibration/drift issues that you have with DC systems--But CT + precision resistors are probably still more accurate).

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • igor1960
    igor1960 Solar Expert Posts: 85 ✭✭✭✭
    Re: Opportunity Load
    BB. wrote: »
    In theory, turning on/off heating elements on AC circuits should not matter. In practice--It can. You may not notice the lights going dim for a faction of second and brightening 5 minutes later when the element turns off. However, cycle the element at 5-10 Hz--A very good chance that everyone will notice lights "blinking" around the home.

    Bill, the problem is that if we use ON/OFF frequency on the heater, then Electric Company might see negative current on SmartMeter, as SmartMeters ADC from what I understand runs at least at 4khz (or maybe even more). So, what I'm thinking is just a dimmer like this: http://www.electricbargainstores.com/product-p/insteon-2476dbk.htm or that: http://www.smarthome.com/2477DH/SwitchLinc-INSTEON-Remote-Control-Dimmer-Dual-Band-High-Wattage-White/p.aspx. I strongly feel heating element is "dimmable"(????)

    BB. wrote: »
    In practice, you only have 0.95 PF for typical "good quality" AC power from a GT inverter--So, ~5% of the time the current flow will be "backwards" relative to the voltage (i.e., complex impedance verses simple resistance).

    Bill, I'm confused here: want to measure current+direction on the main panel and not on GT inverter. As you are saying dirrection might not be that easy to find without voltage -- then probably the best way is to use Neutral/Ground voltage drop/rise to figure out NET current for the whole house, but that would probably require shunt. I was looking for simple non contact solutions.
    Still a little bit confused, why transducer is not changing signal direction on change of AC current, shouldn't magnetic field become opposite?
  • bill von novak
    bill von novak Solar Expert Posts: 891 ✭✭✭✭
    Re: Opportunity Load
    igor1960 wrote: »
    Any ideas on installing element on AC side (not to interfere with MPPT)?

    Pretty hard. You have to accurately (sub-cycle) limit power. If you are doing this to a purely resistive load you can probably do a HF chopper+filter to reduce voltage (and thus power.) However:

    1) This is not an easy design. You need a fast bidirectional switch and a fast control loop.
    2) Any inductive/capacitive load would cause you problems.
    3) EMI problems are likely.
    So, it doesn't switch output polarity on change in AC current direction, Right? Just shows absolute value. Not good. Do you know of any AC current sensor that does switch output polarity?

    When the current is positive, then it gives you a positive voltage at the output. When the current is negative, it gives you a negative voltage at the output. If you look at it with a scope you see a sine wave. But again, this alone is insufficient to tell which direction power is flowing.
    the problem is that if we use ON/OFF frequency on the heater, then Electric Company might see negative current on SmartMeter, as SmartMeters ADC from what I understand runs at least at 4khz (or maybe even more). So, what I'm thinking is just a dimmer like this: http://www.electricbargainstores.com...on-2476dbk.htm or that: http://www.smarthome.com/2477DH/Swit...e-White/p.aspx. I strongly feel heating element is "dimmable"(????)

    Dimmers operate at 120Hz. If the smartmeter has a 4KHz sample rate then your switching frequency has to be higher than that.
    As you are saying dirrection might not be that easy to find without voltage -- then probably the best way is to use Neutral/Ground voltage drop/rise to figure out NET current for the whole house, but that would probably require shunt.

    Same problem there. Knowing the amplitude of current is not enough.
  • igor1960
    igor1960 Solar Expert Posts: 85 ✭✭✭✭
    Re: Opportunity Load
    When the current is positive, then it gives you a positive voltage at the output. When the current is negative, it gives you a negative voltage at the output. If you look at it with a scope you see a sine wave. But again, this alone is insufficient to tell which direction power is flowing.
    Bill, thanx. But I'm now confused even more -- so, it changes polarity, Right? Then why this is not "sufficient to tell which direction power is flowing"? Could you explain, maybe I'm missing something very important!
  • bill von novak
    bill von novak Solar Expert Posts: 891 ✭✭✭✭
    Re: Opportunity Load
    igor1960 wrote: »
    Bill, thanx. But I'm now confused even more -- so, it changes polarity, Right? Then why this is not "sufficient to tell which direction power is flowing"? Could you explain, maybe I'm missing something very important!

    Consider the case where you are looking at the current with an oscilloscope and an accurate current probe.

    If 10 amps of current are flowing into the house you will see a 10 amp sine wave.
    If 10 amps of current are flowing out of the house you will see a 10 amp sine wave.

    How do you tell them apart?
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Re: Opportunity Load
    igor1960 wrote: »
    Bill, the problem is that if we use ON/OFF frequency on the heater, then Electric Company might see negative current on SmartMeter, as SmartMeters ADC from what I understand runs at least at 4khz (or maybe even more). So, what I'm thinking is just a dimmer like this: http://www.electricbargainstores.com/product-p/insteon-2476dbk.htm or that: http://www.smarthome.com/2477DH/SwitchLinc-INSTEON-Remote-Control-Dimmer-Dual-Band-High-Wattage-White/p.aspx. I strongly feel heating element is "dimmable"(????)

    I believe those dimmers would function along these lines:

    http://www.edn.com/design/led/4368149/Efficient-method-for-interfacing-TRIAC-dimmers-and-LEDs-4368149

    Attachment not found.

    So, if you are trying to modulate the heater current to zero out the current to the "smart meter"--It will not work if the meter is measuring PF (i.e., the 4 kHz digital sampling) as meter will "see" the back driven power during part of the cycle...

    That is always the problem with trying to modulate loads to "hide" the GT Inverter from a smart meter. In the US, the utility may just drive down the street (or use google maps) to look for "illegal solar power installs". (in California, even Off Grid solar power system are "illegal" in places connected to utility power--without permits).
    Bill, I'm confused here: want to measure current+direction on the main panel and not on GT inverter. As you are saying direction might not be that easy to find without voltage -- then probably the best way is to use Neutral/Ground voltage drop/rise to figure out NET current for the whole house, but that would probably require shunt. I was looking for simple non contact solutions.

    Yep--Without knowing the voltage (amplitude and timing/phasing) relative to the current (these are vector quantities)--You cannot just connect a current probe and figure out the direction of power transfer.

    Think of peddling a tricycle (peddles fixed to wheel)--Looking from the outside you see legs going up and down. But without knowing the "phasing" of the force vs peddle position, the force could be propelling the trike forward (in-phase, 0 degrees), +/- 90 degrees out of phase (not supplying any forward or reverse average torque), or 180 degrees out of phase and slowing the trike down.

    Regarding contact (shunt, or measuring voltage drop across a length of neutral wire), on non-contact (Current Transformer or Hall Effect Probe). All three methods are pretty much equivalent--Just different methods with differing positive/negative attributes (cost, isolation, sensitive, drift, etc.).

    In general, AC currents are very easy to accurately measure with CT and Hall devices. Shunts are usually used when measuring current with a DC component.
    Still a little bit confuse, why transducer is not changing signal direction on change of AC current, shouldn't magnetic field become opposite?

    With a CT transformer, it is outputting a scaled current. Drop it across a precision resistor, then you have a voltage (that is proportional and in-phase to the monitored current) that is easy to use an Analog to Digital converter to get your digital data stream. Of course, this only "works" with AC current with (typical) a zero DC current bias/frequency below the cutoff frequency of the CT.

    A precision resistor (typically with Kelvin Contacts)--Again, a voltage drop proportional to current. Bring the voltage sense leads to a millivolt A to D of some sort--Again another digital data stream... However, this can do DC because of the direct coupling.

    Hall effect transistors can measure AC or DC fields--But they tend to drift. Since the DC average of AC current is zero--They can be "rezerod" by software averaging (I think--Not my area of expertise). With DC Fields, not so easy (there may be "DC friendly" Hall sensors--again, not my field).

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Re: Opportunity Load

    I should add--AC power is not easy... You can read about the War of Currents between Edison and Tesla.

    War
    of Currents - Wikipedia


    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • igor1960
    igor1960 Solar Expert Posts: 85 ✭✭✭✭
    Re: Opportunity Load
    BB. wrote: »
    A precision resistor (typically with Kelvin Contacts)--Again, a voltage drop proportional to current. Bring the voltage sense leads to a millivolt A to D of some sort--Again another digital data stream... However, this can do DC because of the direct coupling.

    Hall effect transistors can measure AC or DC fields--But they tend to drift. Since the DC average of AC current is zero--They can be "rezerod" by software averaging (I think--Not my area of expertise). With DC Fields, not so easy (there may be "DC friendly" Hall sensors--again, not my field).

    OK, understood now! So, they are not integrating (and I thought they do, stupid me). Thank you guys...

    So, as SmartMeter is used, here is relatively simple solution ZigBee enabled monitor, connected to PC: http://rainforestautomation.com/emu-2
  • igor1960
    igor1960 Solar Expert Posts: 85 ✭✭✭✭
    Re: Opportunity Load
    BB. wrote: »
    I believe those dimmers would function along these lines:

    http://www.edn.com/design/led/4368149/Efficient-method-for-interfacing-TRIAC-dimmers-and-LEDs-4368149

    Attachment not found.

    So, if you are trying to modulate the heater current to zero out the current to the "smart meter"--It will not work if the meter is measuring PF (i.e., the 4 kHz digital sampling) as meter will "see" the back driven power during part of the cycle...

    Bill, again confused here. Assuming LOAD="heating element", how this is different from LOAD="incandescent bulbs"? Are you saying that there is fundamental difference between them?
    And yes SmartMeter (Itron) is measuring PF, by integrating A*V as well as integrating A*V90 (reactive), where V90 is previous 90 degrees early Voltage. Then finding Active by subtracting AV-AV90...
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Re: Opportunity Load
    igor1960 wrote: »
    Bill, again confused here. Assuming LOAD="heating element", how this is different from LOAD="incandescent bulbs"? Are you saying that there is fundamental difference between them?

    For a first order approximation, a filament lamp is a "heating element". For a second order approximation, a tungsten filament is not a bad "current source/sink"... I.e., as you raise the voltage (hotter filament), the resistance goes up, and the current stays, roughly the same. Drop the voltage, the filament cools, resistance goes down, current stays roughly the same.

    So, depending on what your "element" is--You may have some tweak the feedback loop to remain stable.
    And yes SmartMeter (Itron) is measuring PF, by integrating A*V as well as integrating A*V90 (reactive), where V90 is previous 90 degrees early Voltage. Then finding Active by subtracting AV-AV90...

    A modern "smart meter" can measure many things, and if the sampling frequency is 4 kHz, then the Nyquest frequency is 2 kHz or out to the 40th harmonic (if 50 Hz, or 33rd harmonic if 60 Hz power).

    I guess--I keep coming back to the original question... What is it you are trying to do? Trick a "generic" meter (smart or dumb) to not run backwards and cause grief with the billing computers? Or trying to follow the "intent" of the law that there be "NO EXPORT" at fundamental or higher harmonics of power (and any new&improved smart meter which will measure harmonics out to 2 kHz and report back to the utility that there is a sub-cycle net export of power/very poor power factor/very poor derivative of power factor/power quality/etc.).?

    For a utility, they are in the business of selling fundamental 60/50 Hz power. Harmonics are pretty much a 100% loss for the utility (usually shows up as overheating motors and circulating current/overheated transformers).

    Also, utilities really want to charge for kVAH, not kWH. In the US, pretty much all residences are charged for kWH, and businesses (especially larger power users) are charged a poor power factor penalty (basically a kVAH surcharge).

    I could put a "perfect inductor" into a wall outlet and have it pass 10 amps @ 120 VAC.... The kWH meter would see current lagging by 90 degrees and (rightly) charge me zero kWH for that pure inductive load (ignoring the real part of the resistance of the wiring).

    And if it was business--The the utility would see this as a Peak 10a*120v=1,200 VA load... And charge the business based on that load (sort of).

    From the utility's point of view, the wiring, local distribution transformer, back to the transmission lines and back to the power plant--They see the 10 amp inductive load and have to provide 10 amps of current. For the house, they get zero revenue. For the business--The revenue can be so high, that the business will "do something" about the poor power factor (i.e., add capacitors to bring the PF close to 0.95 PF--such as done with large induction motors).

    And why the government mandated CFL lamps are no near as good at saving our infrastructure costs as the government "pretends".

    I buy a 120 watt filament lamp... I pay for the 100 Watts it takes (1 amp * ~120 volts = 120 VA = 100 Watts).

    I buy a 30 watt CFL lamp... I pay for the 30 Watts it uses, but look at the base of the lamp and see that (many) are 0.5 PF... That meas the VA rating for the 30 watt lamp is:

    P=V*A*PF
    VA = P/PF = 30 watts / 0.5 = 50 VA

    So while I reduced my bill to 25%, the utility only saw its RMS current on the line reduced by 50%. For a random CFL light--Not a big deal... Change all the bulbs on a large lighting circuit (or in a whole city), the issue can become quite a problem.

    Harmonics and such with three phase power became a big issue when personal computers (and their power supplies) became heavily used in businesses... The poor power factor (and failure to cancel neutral phase currents) in a large office park would cause building wiring to overheat--Even though the wiring should have been heavy enough to manage the loads.

    There is just a huge amount of issues with AC power where secondary effects can cause problems==From revenue to fires. For a few plug and play GT inverter systems--It is not a big deal. If it becomes popular, then a whole bunch of people/companies/regulators will rain down on P&P/No backfeed installations.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset