On the difference between 'load-flow' and 'phasor' studies
A loadflow (power-flow) simulation is a phasor simulation. It is a phasor simulation of a power system at nominal frequency (50Hz or 60Hz.) It assumes that the system is at sinusoidal steady state and that nothing is changing.
The distinction between a 'load flow' study and a 'phasor study' is that a phasor study can be performed at any arbitrary frequency, say 50Hz, 100Hz, 150Hz, whereas a load-flow study is nearly always performed at the power system nominal frequency (50 or 60Hz.)
The generalised 'phasor study' is useful in the study of power system harmonics, which requires simulation of the power system at 50Hz and its harmonic frequencies 100Hz, 150Hz, 200Hz, 250Hz, ... and so on. This is done by running one separate 'phasor study' for each harmonic frequency of interest.
On the difference between load-flow/phasor and dynamic/transient studies
A load-flow study evaluates steady state operation of a power system. We do load-flow studies to check that elements like transformers, overhead lines, and cables won't be overloaded, and that system voltage regulation is within acceptable limits (-6%, +10% for Australian domestic power supply.)
The time scale of interest is hours to days.
The loadflow study is just an exercise in solving a lot of simultaneous linear equations. There is no time dependent element, no differential equations, or anything exciting. You multiply some big matrices together and that's it.
A dynamic/transient study evaluates the behaviour of the power system when a change occurs. The change could be an increase or decrease in load, a line fault, a change in generator output, or a big motor starting.
The objective is to determine if there will be any detrimental effects on the scale of milliseconds to minutes. Detrimental effects might include - voltage spikes/dips, generator frequency slip, protection relay operation.
A dynamic/transient study must take account of the time-dependent response of the electrical and mechanical parts of the power system.
- Generators and motors have a mechanical inertia
- Capacitors and inductors have energy storage
- Iron-cored transformers have remanence/hysteresis
- Protection relays are digital signal processors which decide whether the power system is healthy or not, based on the history of the signals they see.
- Generators have control systems with sophisticated transfer functions for calculating output voltage set point and governor (throttle) set point
Therefore a transient study involves simulating a system of differential equations evolving over time, with a typical time step of 1 millisecond.
The electrical quantities are still voltages and currents, but there are also a lot of variables in things like 'generator inertial energy' and 'motor rotational speed'.
PS: I do power system studies for a living.
Best Answer
What can you do with phasors?
A phasor is a compact method of writing the important parts of something that varies sinusodially over time. In a phasor you have the information of magnitude and phase, but you omit the information of the actual frequency. Therefore phasor calculations assume that all phasors vary with the same frequency. This is why you only use one frequency variable (\$\omega\$ or \$f\$) in calculations.
So you can describe anything which varies over time in a sinusodial manner with a phasor, as long as you restrict the system to a constant frequency which is equal for all things described with the phasor.
This is not as restricting as it sounds first hand, because from Fourier transformation we know that we can assemble most waveforms by summing up sinusodial waveforms. This is what makes phasor calculations handy: You can apply a transformation to (nearly) any input signal to know of which sinusodials it is composed. Then calculate with phasor calculations the system behaviour at these frequencies and sum up the results. You get the response/output of your system to this input signal. To get the output signal in time domain, you need to inverse transform it.
A graphical help for this are frequency response plots like the bode plot: You see pretty quickly how the system reacts to different frequencies.
What can you NOT describe with phasors?
Anything, where changing the frequency does more than change magnitude and phase. Particularly, this happens if your systems dimensions come near the wavelength of your time-varying signal (then the magnitude depends on the location within the system and the location depends on wave speed and frequency). It also happens if you have dispersion in your system, i.e. the speed at which your signal travels through space depends on the frequency (then you can not add up the magnitudes of all signal at any location because they do not arrive there at the same time). And it also happens if you have nonlinear elements (diodes, transistors,...) in your system - they change the waveform by their nonlinearity and thereby deviate from the sinusodial prerequisite. You need to solve such systems with differential equations in the time domain - not in the frequency domain where phasors are used.
Something that can be solved with phasors, but not right away, is, when you have multiple sources with different waveforms and/or frequencies in your system. In this case you need to calculate the whole input-output-relation separately for each source and sum up the different frequency responses in the end.
So which things can you describe with a phasor?
Everything which can be described by a sinusiodial waveform (this includes cosine) or a sum of these. It is neither restricted to a specific physical domain (electrics, mechanics, acoustics,...), not even to physical quantities at all. You could also describe stock exchange stuff or image processing with phasors if you find a suitable system representation. Phasors are in general a method to solve differential equations in the frequency domain rather than in the time domain - if the differential equations describing the system meet the discussed prerequisites, you can use phasors to solve the equations.