Author Topic: Grid-Tie Principle  (Read 9758 times)

0 Members and 1 Guest are viewing this topic.

jack11

  • Jr. Member
  • **
  • Posts: 95
  • Country: us
Grid-Tie Principle
« on: December 31, 2013, 03:46:44 PM »
I’ve been trying to understand the principle by which a grid-tie inverter pushes the energy into the grid. A good way would be via some mathematical representation from AC circuits, maybe energy/power flow using phasors.

So, can someone explain or quote references, what exactly a grid-tie inverter has to do to push the energy into the grid. Maybe create some small voltage differential, or via some other means. Maybe all this is given in the grid-tie standards from UL and IEEE, I haven’t got my hands on them yet.

The example below may help with this.

Lets say we have two constant-voltage AC sources connected together. Source1 is the grid transformer. Source2 is a RE system inverter.
These two sources are perfectly synchronized in frequency, phase, and voltage waveform. They are of the same type, ex. single-phase, split-phase, etc. All enabling conditions have been validated by source2 to sell energy to source1.
For now assume there are no AC loads attached to source1 and/or source2, just source2 feeding source1.

In practice source1, that’s typically used to supply energy to the loads, is bidirectional and can also act as a sink, that is energy can also flow into it.
Source2 can only act as a source of energy, as you can’t push the energy into the inverter’s AC output.
Since we are dealing with two constant-voltage AC sources, the key thing for energy flow seems to be the direction of the average current flow (if it’s flowing at all).

And so on …………

MaryAlana

  • Hero Member
  • *****
  • Posts: 1200
Re: Grid-Tie Principle
« Reply #1 on: January 01, 2014, 12:52:08 AM »
Not voltage differential, think current differential. If your house is using less current than what your grid tie inverter(GTI) is putting out the rest flows onto the local grid. That reduces demand further upstream and the utility compensates theoretically. In reality your neighbor probably uses the little you generate as excess and the local grid doesn't even see it as current flow back from the house.

OperaHouse

  • Hero Member
  • *****
  • Posts: 1035
  • Country: us
Re: Grid-Tie Principle
« Reply #2 on: January 01, 2014, 07:21:19 AM »
Think of it as a flywheel that is kicked occasionally to keep it spinning. You are thinking of this elaborate phasing.  Most of these are like MSW inverters, just dumping as much current as the can into the line.   Once the grid tie sees something like 60V it just dumps in as much current as it can.  The inverter can easily adjust current tens of thousands times a second.  Like saussage, you don't want to see how it is made.

Flux

  • Super Hero Member Plus
  • *******
  • Posts: 6275
Re: Grid-Tie Principle
« Reply #3 on: January 01, 2014, 09:07:40 AM »
Another way to think of it.

Regard the grid as infinite, nothing your inverter will do to alter it. If both voltages are the same nothing happens but as soon as your inverter volts exceed the grid current will flow into the grid. this current will be limited by the source impedance of the inverter supply and will alow the available current from your wind or solar to feed the grid.

Normally you need to match the available power to keep your panels or wind turbine on optimum production and that is where the inverter curve adjusts the power to suit the load to get the best match. It does this by altering the voltage ratio of the source and the grid.

Flux

jack11

  • Jr. Member
  • **
  • Posts: 95
  • Country: us
Re: Grid-Tie Principle
« Reply #4 on: January 02, 2014, 02:34:17 PM »
ok, thank you all.

thinking about it some more, no specific voltage differential seems needed, just a distributed power source sitting on the grid, matching it's voltage waveform, and ready to pump current into the grid if the grid loads somewhere down the line demand it.

This brings up another question, about the priority of who gets to use what power.
Say, I now have some local loads in the house that's equipped with this grid-tie system.
If a local load turns on, is it guaranteed to use the energy from the local grid-tie RE system first, or from the grid first, or half and half, or what?
If it uses the energy from the local grid-tie system first, and then uses the grid to supply the overage only, then what is the mechanism that allows it to prioritize usage this way ???
This has a big impact on the financial benefits from a grid-tie system, in case of the unsymmetrical utility rate structure (where they pay you much less per kWh than you have to pay them).

Still. it would be nice to have some math model for it all, and be able to show via a simple computer simulation that it works the way you think it does. I just need to see that sausage being made to convince myself I can eat it ;-)

Frank S

  • Hero Member
  • *****
  • Posts: 1622
  • Country: us
Re: Grid-Tie Principle
« Reply #5 on: January 02, 2014, 07:52:08 PM »
Jack think of a grid tie as a pool of water fed by a stream and a well and several dipping their cups in at the same time.
 the stream is constantly filling the pool the well occasionally adds to this once the water is in the pool there is no way to tell which cup is filled with what water.
 if you are pumping from your well and filling your cup at the same time you still get what is available. Be it from someone else's well or the grid it is all the same the only way to say I want to drink my own water only is to  disconnect from the grid.
 you can however be consoled with the knowledge that sat least some of the water or in this case electricity, was yours 
I live so far outside of the box, when I die they will stretch my carcass over the coffin

boB

  • Sr. Member
  • ****
  • Posts: 383
  • Country: us
    • boB
Re: Grid-Tie Principle
« Reply #6 on: January 02, 2014, 09:12:43 PM »
It does this by altering the voltage ratio of the source and the grid.

Which in turn determines the direction of power and how much.

Note the words "voltage ratio"...   That is the key.

Similar to a variac on the AC side of the inverter.

It's really just a simple variation of Ohms law... Power and sign and magnitude of current,
resistance of the grid, etc.  If the grid impedance is high, the grid tie inverter selling back
will tend to raise the grid voltage....  And the opposite, too... Just as a high impedance
(weak) grid voltage tends to sag when loaded down a lot.

boB
« Last Edit: January 02, 2014, 09:18:32 PM by boB »

OperaHouse

  • Hero Member
  • *****
  • Posts: 1035
  • Country: us
Re: Grid-Tie Principle
« Reply #7 on: January 03, 2014, 02:37:01 AM »
Just remember, the sum of the currents at any node is always zero.

Caleb

  • Newbie
  • *
  • Posts: 20
  • Country: us
Re: Grid-Tie Principle
« Reply #8 on: January 03, 2014, 07:01:35 AM »
This link was posted at The Back Shed.  I can't speak to how well it works, but it might be instructive.

http://www.thebackshed.com/forum/forum_posts.asp?TID=2103

jack11

  • Jr. Member
  • **
  • Posts: 95
  • Country: us
Re: Grid-Tie Principle
« Reply #9 on: January 03, 2014, 03:04:12 PM »
ok, so there is some sort of balancing act a grid-tie inverter does with the grid, by adjusting the ratio of its output AC voltage to the grid's voltage. That sounds like a voltage adjustment an off-grid inverter could also do depending on its load, except that typical inverter loads are passive, and the grid is an active load which behaves quite differently. I scanned thru the PhD dissertation Caleb provided, it does seem to have some explanation of this, but a lot of reading too.

Regarding the priority of use, Frank is right, I wouldn't care whose energy I use. Except that the utility meter is between the grid and the local grid-tie RE system/local loads. Given the asymmetrical rate structure, if I put my local energy into the grid first, and then use some grid energy to power my local loads, then I get cheated (my sell rate is much less than my buy rate). On the other hand, if I somehow force my local energy into my local loads first, and only sell the excess energy to the grid, then I don't have to buy from the grid at all, and at least I get something for what I sell. This of course would not be an issue if the sell and buy rates were symmetric. All this is assuming that the utility meter is in the place I think it is.

Caleb

  • Newbie
  • *
  • Posts: 20
  • Country: us
Re: Grid-Tie Principle
« Reply #10 on: January 03, 2014, 03:30:20 PM »
I vaguely recall someone referring to grid tie as "injecting current" into the grid.  If you have a boost converter, it really doesn't care so much what the output voltage is.  The voltage will rise until current flows out of the inductor into the load.  Perhaps a similar principle is applied to grid tie.

dnix71

  • SuperHero Member
  • ******
  • Posts: 2250
Re: Grid-Tie Principle
« Reply #11 on: January 03, 2014, 04:20:28 PM »
jack11 is wondering if it matters if his RE injection is on his side of the meter or the utility side. It should be at the meter itself in which case it wouldn't matter.

The injection can't occur beyond the house meter, unless there are 2 meters, otherwise you would be giving the utility free electricity and then buying it from them. It could be inside, between the meter and the house main panel, that might be the easiest place to install it, but it should be injected at the point of service as an extra set of wires to the meter itself.

The illegal backfeed inverters sold on fleabay (originally diverted from legal use in Holland) inject inside the house, so as long as the injection is less than or equal to the whole house consumption, the meter never runs backwards and the utility company wouldn't know what you were doing.

Rob Beckers

  • Full Member
  • ***
  • Posts: 130
  • Country: ca
Re: Grid-Tie Principle
« Reply #12 on: January 06, 2014, 05:01:07 PM »
In an attempt to make this thread a little less vague:

There are two types of inverters, one is used for off-grid, turning battery DC into local grid AC. The other is the topic of this thread, a grid-tie inverter that takes DC (from PV, wind, or hydro) and turns that into grid AC. The two may seem similar, but have a very fundamental difference.

The first, the off-grid inverter that works off batteries, creates the Voltage and frequency, it controls it. It tries very hard to keep the grid Voltage that it produces constant, much like a battery would. It functions as a Voltage source.

The second, the grid-tie inverter, has its micro-controller set up to follow the grid Voltage and frequency. It does not make it, it merely follows what's there. It tries very hard to keep its output current constant (as dictated by the MPPT algorithm in there, trying to get the most out of the DC source). As such it doesn't 'care' what the Voltage is (within limits), and if it needs to raise its output Voltage a little to keep that current flowing it will do so.

The grid-tie inverter normally raises the grid Voltage (to be able to pump that current into the grid). Usually just a little, sometimes a lot; I've seen sites where pushing 10kW into the grid raised it by 15 Volt. It all depends on the resistance/impedance that inverter 'sees' looking into the grid.

Where that current (and power) goes is of no concern to the inverter; much is likely to be used locally, since that is simply the lowest resistance path. How it's distributed between your house, the neighbours, or the rest of the grid is entirely a matter of Ohm's and Kirchoff's laws (much like several water pumps supplying a set of hoses; where the water goes depends on the resistance to flow in each of the directions).

The hardware for off-grid (Voltage source) and grid-tie (current source) inverters is pretty much the same. The difference is almost entirely in the programming of the micro-controller that runs the show.

-RoB-

MaryAlana

  • Hero Member
  • *****
  • Posts: 1200
Re: Grid-Tie Principle
« Reply #13 on: January 07, 2014, 02:09:07 PM »
And when you think of how much power even a small town like mine with 120 houses is using that little bit of energy you are adding to the grid might equal the line losses in your neighborhood

jack11

  • Jr. Member
  • **
  • Posts: 95
  • Country: us
Re: Grid-Tie Principle
« Reply #14 on: January 07, 2014, 02:46:50 PM »
ok, I think I am homing in on better understanding.

The last posts from Caleb and Rob, and earlier from others, are pretty much in line with what that PhD dissertation says.
I made an error in assuming that a grid-tie inverter functions as a voltage source (like a off-grid inverter). In fact, there would be a problem with this (see the dissert), and it apparently needs to function as a current source.

I think the dissert says exactly as Rob explained, once the inverter synchronizes and validates the grid waveform, it works off a current feedback sensor to control its output current to follow some reference profile the dissert calls "demand current" (see the inverter's ATP model code in dissert - fig A.5).

One thing that's still unclear to me is, how is this "demand current" reference derived by the inverter to follow?
Is it simply the max amount of current the inverter can produce within its own limitations and under the current grid/load conditions (voltage, impedance), or in some other way?
What variables (analog or digital) are used to create this "demand current" reference, and how often does it need to be recalculated?

Rob Beckers

  • Full Member
  • ***
  • Posts: 130
  • Country: ca
Re: Grid-Tie Principle
« Reply #15 on: January 07, 2014, 05:11:31 PM »
Jack, the current is set by the MPPT algorithm in the inverter, and therefore differs a bit if you're talking about a PV inverter or a wind/hydro inverter. To stick with solar PV (for simplicity), the MPPT algorithm figures out what the best load is on the solar panels for the most energy extraction (it can do this by going from no load to a high load very quickly, a sweep, or once it has a ballpark MPPT point it can do this through small variations in the load). Either way, the micro-controller ends up with an operating point for the solar panels that specifies the current drawn from them (and therefore the Voltage).

Depending on the circuitry of the grid-tie inverter the input DC voltage can be lower or higher than the grid voltage, so there can be a boost or a buck stage that takes care of the Voltage conversion (and the output stage, normally a set of 4 electronic switches such as IGBTs, or MOSFETs, also called an "H-bridge" because the circuit diagram for this looks like the letter "H", can also take care of lowering the Voltage by behaving like a buck stage). The output stage, or H-bridge, will do the conversion from DC into AC.

Now to get back to the current it needs to draw from the source for the MPPT operating point: The micro-controller will either have a current sensor on the input side and regulate the output in a feedback loop such that it meets that input current, or there's a sensor on the output side and it uses math to convert the input current into the output current it needs to draw to meet that operating point, again in a feedback loop. The feedback loop varies the output Voltage such that the current going out satisfies the operating point.

Of course there are limitations, there is a maximum current that those switches can safely handle, and the micro-controller will take that into account. If it bumps into those limits (Voltage or current) it will limit the output and fall out of MPPT. Grid Voltage is somewhat secondary to this; the inverter will raise the output Voltage as needed to meet the current operation point. Again, up to certain limits, set by UL1741; if the Voltage dips above or below regulatory limits the micro-controller has to disconnect from the grid within a specific time (depending on how far it's over or below those limits).

Hope this helps!

-RoB-

DamonHD

  • Administrator
  • SuperHero Member
  • *****
  • Posts: 3618
  • Country: gb
    • Earth Notes
Re: Grid-Tie Principle
« Reply #16 on: January 07, 2014, 11:43:28 PM »
Thanks for that very clear explanation.

I pinned this thread earlier because this is excellent how-does-it-work info!

Rgds

Damon

jack11

  • Jr. Member
  • **
  • Posts: 95
  • Country: us
Re: Grid-Tie Principle
« Reply #17 on: January 09, 2014, 02:51:45 PM »
thanks Rob, I think this explains the general principle.

It looks like that "demand current" reference I was looking for is simply the PV array's Imp from the MPPT algorithm, the inverter needs to track using its AC output as control.

I assume the "input current" you refer to is the PV array DC current going into the inverter's DC-DC stage. Also, the "output current" is the inverter's AC current going into the grid from its DC-AC stage. Either current can be sensed using an appropriate current sensor. Then the control variable used to maintain input or output current tracking is the inverter's output AC voltage. Please correct me on any of this, I'd eventually like to write a simple computer simulation to model this scheme.

So, if we use an input current sensor, then we track the input current directly. If we use an output current sensor, then we need to synthesize the input current sensor using that "math" you referred to. Can you give any insight into this math, or at least the inverter parameters that need to be modeled to do that output-input current conversion?

jack11

  • Jr. Member
  • **
  • Posts: 95
  • Country: us
Re: Grid-Tie Principle
« Reply #18 on: March 03, 2014, 01:16:04 PM »
for information, this is another description how a grid-tie inverter functions, this one taken from the tigoenergy.com website.

“TRIAL AND ERROR” MAXIMUM POWER POINT (MPP) TRACKING

By observing the topology of most installations today, the most widely accepted approach for cost and reliability is to have a central inverter with a variable DC input from the array. The inverter performs the DC to AC conversion necessary to deposit energy production onto the grid. These single or multi-stage conversion processes from the leading inverter manufacturers (DC/DC step-up for isolation and DC/AC) have been optimized over 50+ years, are highly efficient and well accepted by global regulatory bodies and power companies.

The MPP tracker within the inverter attempts to keep the array (or string) at the highest power output pos-sible. To find the point at which the entire system can produce the maximum power at the current solar irradiance point, the tracker usually applies a “trial & error” algorithm which adjusts its current draw (load) on the system. By measuring the new DC power input, the tracker will determine whether to continue the adjustment in the same direction or reverse course. This process is constantly looking for the peak power point but rarely finds the system working at this point (only instantaneously during transitions). There are many variants of this algorithm but with input data limited to system DC voltage and current, all have limited accuracy. The task be-comes significantly more complex during times of changing irradiance (ex. cloud cover, shad-ing) as each module’s maximum power point is dynamically moving. System stabilization may take several minutes after a cloud has passed. Because each module has a series of by-pass diodes, a significantly under-performing module can be “turned off” when the current drawn from the inverter exceeds its ability to provide power.


boB

  • Sr. Member
  • ****
  • Posts: 383
  • Country: us
    • boB
Re: Grid-Tie Principle
« Reply #19 on: March 03, 2014, 02:02:41 PM »

Trial and Error may also be perturb and observe.   There is no official name for an algorithm called trial and error that I have seen but that pretty much does sum it up correctly.  The best that an algorithm can do is to spend as little time as possible away from the MPP voltage while it is trying.

Also, a solar "farm"  (as opposed to a residence) will not have any partial shading and will not have to spend a lot of time too far away from the max power point voltage...   At least not so far away that it has to find a secondary maximum MPP V that you have with partial shading.  Of course, you can have a bad module once in a while but that shouldn't affect the whole thing too much.

  Adding modular MPP optimizer modules in a system without partial shading (i.e. with a central inverter) will (typically) just reduce the output power due to extra losses.   They do make sense for systems with partial shading though.

boB