The basic design of stand-alone inverter is totally different from the grid-tie inverter. In stand-alone inverters the loading on the system is based on the available and forecasted generation from RES generation systems. one should have a clear picture of generation (forecast) before loading the system.  

As the power generated by the RES cannot be controlled as the load demand changes because the generation is totally depend on the environmental conditions. Keeping this in mind you have to estimate the maximum load demand as it is stand-alone system, the system will be designed for a particular load area or centre. So, it will be easy to design the inverter controller. A lot of literature is available regarding the control of inverters.

Considering the worst case condition for load in the question you have to install a huge reserves for energy storage as you are not sure about the loading of the system. which is not a good design of stand-alone system.

Hope this may help you.... Best of luck 

@Pavan Thanks for briefing about the overall system design of Standalone systems. The answer addresses part of my question, but I'm still not clear how the control system design is done. Ex. I have a 1kVA standalone inverter with sufficient power backup that can support full loading of the inverter, and I turn on a 900W load (resistive for ex.) at say the peak of the Voltage waveform, how would the load prediction be done and the distortion created by the load on the Voltage Waveform be mitigated?

Dear Seetharaman,

in stand-alone condition, your inverter remains a basic voltage source.

As long as your primary source generates sufficient power, providing the voltage to the consumers is good enough.

All you need to do is provide the necessary control to generate the output voltage and at the same time monitor and if needed limit the output current.

As long as the change in load remains within the specified operation area of the inverter's power stage, the world is fine.

Current has to remain the degree of freedom to not overdetermine the system.

How quickly the system can follow a step in load depends on the source more than on the inverter.

Once you reach the maximum current your control has to decide whether to reduce the voltage to keep the current below maximum - usually not an option - or to turn off with an error message.

Reducing the voltage typically is not an option as the consumers connected within your island may internally comensate by drawing higher currents, especially those that feature switch-mode power supplies with a wide input voltage range.

Cheers - Martin

@Martin Schulz Thanks. Completely makes sense. 

Does it matter to the end equipment if there is a small glitch(lasting less than 5ms) in the voltage waveform during step load transition, as long as the glitch falls within the equipment's rated voltage? What I'm asking is that, would this small glitch cause damage by any chance? 

I remember reading that all equipments are designed to withstand 3 cycles of voltage surge/sag, is this true?

Dear Seetharaman,

this depends on the type of equipment - I don't know about a general capability regarding a so-called low-voltage ride through.

Again - anything that is powered by SMPS will not get into trouble as the local capacitor will provide the energy buffering.

Light bulbs might flicker but this is not important.

As today most consumers have a local power converter, I suppose a glitch of 5ms that goes down to no less than 50% of the rated voltage will not get anything into trouble.

Cheers - Martin

It important to control the output voltage and power output in order to avoid frequency deviation by using dq control