With domestic and industrial battery backed solar power in system the allowable imbalance have increased tremendously, only practically could be given with areas, as working power factor could be near unity(lag)….

As rightly pointed out while placing the question, inertia is definitely an important factor to make it fast or slow during imbalance (so far as change in angle is concerned), overall damping too play an important role. Hence power factor is an issue. Better power factor on account of resistive load incident in the system mostly gives better damping obviously. Now let some practical experience be shared. Very often (statistically may be checked) angular instability has occurred during light-load condition more than when the system is around peak-load. People very often say due to attentiveness of operators it happens (as if due to casualness or everything taken granted during light-load condition it is so). But it's also a fact that during light-load with overall system power factor too deteriorating usually (as not the load alone, but the system too contributing) damping reduces, allowing for larger excursion of angle during transient lasting for more time. Regarding quantification in terms of MW or in percentage of unbalance it is difficult to say. As one may see from the above-mentioned experience, possibly and rightly it will differ in the extremities of peak-load and light-load. That's why in practice while operating the grid, there has been a use of Figure of Merit or in real terms MW per Hz for the system. Larger it is system is more secured. Again, it is also a measure of damping indirectly, vis-à-vis inertia.

Thanks for all researchers for their answers.

@Subrata Mukhopathaya I agree that there will no be specific quatification or even a range for such imbalance. I just would liked to make sure of that. Thank you so much.

Let us hear further through sharing of practical experience while operating the grid or through numerical computation for a sizeable system at least.

Dear Sir, as far as the frequency stability is concerned a commonly used performance criterium states that the loss of the largest generator of the system shall not produce a load shedding. This is an aspect of the well known N-1 rule. More complex situations can be considered e.g. the loss of a full pole of a HVDC link, the loss of an interconnection tie, a bus-bar fault and so on. In case of multiple contingencies or very unlikely events, a frequently used performance criterium is that the considered event shall not cause the disruption of the system even if a portion of the load is lost. The maximum amount of the power unbalance a system can withstand depends on several factors such as system inertia, response speed of the governors, reactive power margins, load response to the variation of the frequency and of the voltage, the available reserve for the primary frequency regulation (the latter value is generally fixed considering a credible power unbalances). From what above the possible power unbalance is a characteristic of each power system. In practical terms unbalances of say 5% of the load should be easily withstood, unbalances above say 10% - 15% could be very hard.

Best regards, Giulio Santagostino

Whatever has been told by Giulio is very much appreciated due to clarity to the point. Further if one tries to see power angle stability, both from small disturbance and large disturbance perspective considering the system as highly nonlinear, the degree or margin of stability will depend on point of system operation. Yes, by this point, it is meant again the overall loading of the system. At some point, usually lower loading, margin will be available much more than while working full to the brim.

Allow me to para phrase the question; How much deviation in frequency ( steady state or dynamic) and for how long can the deviation be tolerated or allowed as per the code/standard of any specific power system. The largest credible contingency , the inertia , K , Frequency response characteristic , first stage of defence UFR, P,S T reserves etc etc would influence the number ; In Indian interconnect , reference contingency is taken as 4500 MW, Frequency deviation of 0.05 Hz is allowed ; FRC ranges from 15 GW-25 GW per Hz ; Hope this helps ;

Dear Omar,

As it is rightly answered by the eminent experts of the field. I just want to add some frequency security parameters followed by interconnections.

The conventional frequency response requirement range prescribed by the ENTSO-E for dead band, droop and response delivery time is 0-500 mHz, 2-12% and 2 s-30 s, respectively. The standard technical requirements for new fast frequency reserve services by the ENTSO-E comply with the reserve activation instant at zero seconds immediately after an imbalance. The maximum time required for a complete response activation is 0.70 s for a 0.5 Hz frequency deviation from the nominal value, 1 s for a 0.4 Hz frequency deviation, and 1.30 s for 0.3 Hz frequency deviation. The minimum response for the short duration of support is 5 s and 30 s for the long duration of support.

North American Electric Reliability Corporation (NERC) and Federal Energy Regulatory Commission NERC BAL-003-1, FERC Order 842 specifies the criterion for dead band 36 mHz, for droop 5%, and the delivery should be without any delay and sustain for at least 30 s.

Newly admitted fast-responding frequency response services in the global power grids such as Synchronous Inertial Response (SIR), Fast Frequency Response (FFR), Enhanced Frequency Response (EFR) and Dynamic Regulation Signal (RegD). The SIR service is provided instantly as compared to the FFR, EFR and RegD services, which are provided within 0-2 s. The deadband for the SIR, FFR, and EFR services is 17-36$ mHz, 15- 200 mHz, 15- 50 mHz, respectively. The droop for SIR service is 3-5 % and EFR service is provided through two envelopes.