Yes thanks Ken. I still don't really understand but I am now sure it will work as advertised On 25 Jun 2016 6:57 a.m., "Brent Brown" wrote: > Thanks Russell, thanks Ken. Good insights on the workings of grid tied > inverters. > The picture is much clearer now, to me anyway. > > RussellMc wrote: > > > My friend Ken has useful experience in the interface aspects of grid ti= e > > inverters. > > I sent him a portion of the discussions here and he provided the > following > > comments: > > > > > > Russell, > > > > None of this is likely to be "news" to you, but some of your audience m= ay > > find it helpful. > > > > The principle at the core of any grid-tied inverter is that it acts as = an > > AC current source with a voltage compliance that exceeds the peak volta= ge > > of the AC mains. > > > > Once you have that, and a means to synchronise the current injection > > waveform from the inverter with the AC mains voltage, you can happily > > deliver current (in any phase relationship to the voltage) into > > > > the grid supply. > > > > The fact that the grid has a very low impedance makes the job easier > > because the process of injecting current does not significantly perturb > the > > voltage waveform - thus making continued synchronisation reliable. > Detection > > of any such voltage perturbation should it occur is one means by > > which grid-tied inverters detect islanding. > > > > Well-designed grid-tied inverters can also be set up to work with a > > relatively "soft" AC supply. For instance you can use an SMA Sunny > Island > > inverter to establish a "local" grid and then add Sunny Boy (or Windy > Boy) > > grid-tied inverters to inject further power in order to supplement that > > grid supply. > > > > The basic principle has been around more or less since the beginning of > AC > > power distribution - but in early times took the form of synchronous > > rotary machines that delivered reactive power (current out of phase wit= h > > the voltage) to the supply in order to effect power factor correction. > > > > Note that for the purposes of many grid-tied inverters an AC current > source > > is just a DC current source (usually implemented as a high-frequency > > switch-mode power converter - often in boost topology) plus a means of > > commutation (often a full bridge operating at the supply frequency of 5= 0 > or > > 60 Hz) to deal with the voltage reversals of the AC grid supply. In > effect > > the commutation process takes a half-wave "rectified" output generated = by > > the current source and converts it to a full-wave AC output compatible > with > > the (sinusoidal) grid supply. > > > > The issue of drawing current from the grid while simultaneously injecti= ng > > current into the grid is easiest to understand if you just regard the > grid > > as an AC bus capable of absorbing or delivering any amount of power you > > like. Ignoring the added complexity of reactive power (when the curren= t > > and voltage are not in phase), if you draw more current than you inject > net > > power is delivered to you, and if you draw less current than you inject > net > > power is delivered to the grid. > > > > In the real world things get a bit more hazy because (in this part of t= he > > world at least) utility companies pay a lot less for energy (kWhrs) you > > inject into the grid than they charge for energy that you take from the > > grid. They will often use separate energy meters - one operating > > conventionally to record the energy consumed by your electrical load an= d > > the other connected "backwards" between your own source of electrical > > energy (PV panels, wind turbine, micro-hydro, etc.) and the incoming gr= id > > supply. The AC bus still exists (on the grid side of the two meters), > > but your electrical load and your own electrical energy source are not > > directly connected. Many modern digital energy meters can combine the > > separate measurement functions into one device - in which case the gri= d > > connects to one port of the meter, and both your load and your own ener= gy > > source (connected in parallel) connect to the other. Then, by monitori= ng > > the phase relationship between the voltage and the current flowing > through > > the meter, energy consumed and energy delivered can be separately > metered. > > > > As an interesting experiment that will almost certainly lead to an > > epiphany, take two suitable identical iron-cored transformers. Supply > one > > primary directly from the mains and the other primary via a Variac. Se= t > > the Variac output to match the mains voltage. Connect the two > secondaries > > together via a low-resistance current shunt - taking care to get the > > phasing the same. Use an oscilloscope to display the secondary voltage > on > > the transformer fed from the Variac, and the voltage across the current > > shunt. Adjust the Variac voltage up and down slowly while watching the > > waveforms. You can alternatively conduct the experiment in simulation = - > > using LTSPICE or similar. > > > > There is a widespread misapprehension that loss of synchronisation > between > > a grid tied inverter and the grid supply will necessarily result in > > destruction of the inverter. While that may be the case for a > > poorly-executed design, the fact that the inverter is a current source > > means that it is potentially capable of driving a controlled current in= to > > the grid supply regardless of the instantaneous voltage of that supply > > (within the normal limits of the peak AC voltage). So for instance if > the > > inverter has an internal DC bus (often called a DC link) of say 600V, i= t > > can deliver current into a 230VAC (rms) supply when the supply voltage = is > > at its positive peak (+325V) and the voltage difference is 275V, or whe= n > > the supply voltage is at its negative peak (-325V) and the voltage > > difference is 925V. Not all inverters are designed to deliver current > over > > the full range of voltage difference (because that's not typically > > necessary for most applications) - but it is certainly technically > > possible. > > > > In the past, most grid-tied inverters have been unable to accept > > significant power from the grid (i.e. anything greater that what they > need > > for their own internal "housekeeping"), but there is a class of grid-ti= ed > > inverters that can have an associated battery bank - and these (often > > referred to as inverter-chargers) can take significant energy from the > grid > > and deliver it in order to charge the storage batteries. > > Inverter-chargers will become increasingly common as battery technology > > improves and becomes more cost-effective and systems like Tesla's > > "PowerWall" which support temporal "load-shifting" become popular. > > > > -- > http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive > View/change your membership options at > http://mailman.mit.edu/mailman/listinfo/piclist > -- http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist .