In grid synchronous machines are connected, practically, grid cannot be operated with, unity power factor, the zero reactive power( as it is a steady state stability limit, only theoretical), transient stability, with best power factor would be 0.8 lag, and higher would be at the cost of stability.

Dear Amit,

the control scheme seems to be fine. If the AC and DC voltages have been set at adequate values, the reason for not being able to control Q might be a deficient tuning of the controllers. Have you already considered that possibility? Are you experiencing problems with the control of the DC bus voltage as well?

Best regards,

Raúl.

In the grid connected inverter system usually we will use SPWM or space vector pulse width modulation. Hence, even though we force reactive power to zero through dq control, little amount of harmonics will be present in the inverter output. This will lead the small amount of reactive power requirement  in the grid. So that, the grid reactive power will not be zero.

One can find a solution as reactive energy to be zero only on fundamental harmonic, in static, but is impossible reactive energy exchange to be zero on the higher harmonics.

can the value of X/R ratio be one of the reason??

The reactive power can't be zero if the tuning of the P.I controllers is missing and also there may be some inductive components which may lead to a small value of Q rather than absolute "0" value. 

Dear Mr. Roy, there are two different issues, although interconnected to each other. The first is what Mr. Desai pointed out, i.e. the mandatory reactive power demanded for the operation of a synchronous generator. Of course all ractive components of a grid play a role too, but the main thing is the required reactive power for the generator. So, you are obliged to have even a small amount of it (pf 0,8 usually). The other issue is the possibility of really having a zero reactive power, despite the synchronous generator and stability issue. The answer is again NO, because some reactive power will always be produced by the higher harmonics and the distortion from your power electronic components as most of our colleagues pointed out. Therefore you shouldnt and you cant! You shouldnt because of the generator and you cant because of the inverter. 

@ Raúl Sarrias-Mena: DC bus voltage is controlled, but the value of reactive power comes to be negative...in my simulations, when I m measuring it at the point of common coupling after the LC filter.

Hi Amit,

if you are modeling the grid as an ideal AC source (which I had assumed in my previous comment), the grid will behave as an infinite bus, and thus you should be able to control the reactive power exchange with the grid. Besides, in the reference paper the authors state "the reference for the reactive current is usually set to zero, if the reactive power control is not allowed. In the case that the reactive power has to be controlled, a reactive power reference must be imposed to the system." Therefore, they indicate the ability of the control system to regulate the reactive power exchange with the grid, and even to set it to zero.

Moreover, I cannot appreciate if a negative reactive power means consumption or generation to grid. I would suggest that you measure the reactive power at different points of your system, and observe how it varies. Thus, you will be able to figure out the contribution of all the elements in your scheme to the reactive power exchange with the grid.

The Explanation given by A.A.P. Pareskevopoulus seems to be Plausible.

We have to bear the Following points in mind.

Switching Harmonics will be present and we try to control and accordingly Reactive Power Control and Improvement of Voltage Profile takes place.

There is always present the Resistance(Including Armature Reaction Component and the Inherent Reactance of the Synchronous Machines.If this Reactance Component of the Machines are also Neutralized we may approximate to D.C. Generation and Unity Power Factor. We try to do it for getting Zero Reactive Power.

But the Fundamental Principle of Control Systems Requires that there should always be an error,however small it may be,between the controlled output and its Reference Value for the Control System to Function.If we get the Ideal situation of Zero Reactive Power there cannot be any Control Action.It is only an Ideal Picture and not Practical.

P.S.

I may also add that there will always be a Time lag,however small it may be, between the error and the application of corresponding correction to the system,even it it be digital/online control,due to the inertia of the Control System Components and also the time taken to Execute the Corresponding Program.

That is why we say that in Real time Control Systems,the Algorithm used should be very fast though not very very Accurate(But should be Admissible) because by the Time of Application of Correction the Controlled Parameters Might change even it it be a very Slight Change, in any Dynamic System.

This Aspect also may explain your Problem.

P.S.

According to my opinion, not to pump the reactive power to the network is reduction the currents through the power electronic switches in grid-connected inverter. Because:

 The value of the harmonic power pumps to network in a tied-grid inverter is always constant (because when the pumped power is increased to network, THD will decrease). Therefore, if the active power is constant, THD will decrease when we pump the Q to network. So, to maximize the capability of carrying current of electronic switches, the zero of Q value will be selected in a tied-grid inverter (PF = 1.0). However, We may control the Q to network when necessary. As a demostration, the inverter that pumps the Q to network only is STATCOM device. The statcom has a structure like a grid-connected inverter

In a grid-connected inverter, you always have a filter (often an LCL) at the output ... if you inject pure active power in the grid, the inverters must still absorb some reactive power because of presence of the filter capacitor. Conversely, if your inverter supplies pure active power, the grid is supplied with some reactive power... so, the reactive power delivered/absorbed cannot be zero for both inverter and grid simultaneously in the presence of such a filter ...

Many important things to check has been already mentioned here such as delays due to measurement and control algorithm execution. Naturally it is important to check that your measurements are OK as voltage and current ripple may not be correctly taken care of if your sampling is not frequent enough. If you are using filters to remove the ripple (as you should) please note that they affect the phase angle of the fundamental as well. Further, check the operation of the PLL as it may not be locked in correct synchronism with the fundamental. Naturally the PWM of the inverter cannot produce higher voltages than the DC voltage allows, so the operation of the modulator is good to check as well.

Hello,

Have you solved this problem？ I met same problem as you. The reactive power of inverter is not zero although I have locked it to zero. Could you tell me how to solve this problem? Thank you.

me too

did u get the answer, how to make reactive power control