Thursday, January 17, 2013

Grid Disturbance

Java Bali grid in the morning has total load 18.500MW with total capacity 19,990MW. It was normal status due to the biggest power plan is 815 W (Normal status if Total Load 18,500MW plus biggest power plant 815MW is lower than Total Capacity 19,990MW).

Some how on 500kV east Java sub station got protection earth fault. It makes power supply from east java around 3000MW rejected. It makes the Java Bali grid status has changed from Normal to Deficit 1,500MW due to total capacity (19,900 minus 3,000MW ) is lower than total demand (18,500MW).


















The frequency grid drop from 50Hz t0 48.5Hz. During frequency on 48.5Hz, the voltage on central Java drop to from 518kV to 496kV, Load shedding action plan had carried out.


In the panic situation, operator only concern about frequency grid during load shedding because if frequency continuous drop to 48.3Hz, island operation created.

This time was lucky, the frequency stood still on 48.5Hz (action plan step 6th done) but the voltage on central Java spike from 496kV to 535kV. The operator should be reduced MVAR setting while this happen. But it was too late. One Circuit Breaker got protection open due to the voltage reach 535kV, but it can normalize again in the short time.

To help the grid when disturbance happen, each power station have two regulation.

1. Primary regulation, Each power station must set on Free Governor Mode. It means while the frequency still inside dead-band(49.7Hz - 50.3Hz), the turbine control valve still stand still. If it already outside dead-band, turbine control valve will response VERY FAST.

2. Secondary regulation, LFC mode.
Load Frequency Control Mode for restore the grid back to 50Hz SLOWLY

Most power plant loads vary in an irregular manner. They can usually anticipates the hour to hour changes but not their exact values. The fuel and air feed to boiler furnace must be varied accordingly or steam pressure and turbine shaft speed will get out of hand.

At steady load, energy to plant input exactly balances to plant output and losses. At the same time a bank of light is switched on the system, this reduce the over-all electric resistance of the system and a greater current flows through the generator armature and connected circuit. Higher amperage raises the armature magnetic field strength, increase the turning resistance offered the field rotor, this slows the generator and turbine shaft speed. By giving up some of its rotating kinetic energy the shaft momentarily support the added load.

FREE GOVERNOR MODE
When frequency changes in the grid every Turbine-Generator unit reacts and adjusts its generation as dictated by power frequency or droop characteristic. For instance when frequency falls below 0.1 %.

Generation has to be increased 5 % (FREQUENCY BIAS ACTIVATED) with droop also 5 %. In Indian situation most of the generating units operate at their peak values and no additional generation is possible. With the result many units do not increase their generation and load shedding is resorted to.

Usually setting dead band for Free Governor Mode 49.7 Hz to 50.3 Hz.
Setting for frequency bias 49 Hz to 51 Hz.


LOAD LIMIT MODE
In some cases, due to various operational reasons generating companies do not like to their machines to respond, even though spare capacity is available. The governing is bypassed. If most of the generating stations in a grid do not respond naturally, there is a danger of grid becoming unstable also.

In the recently approved Grid code it has been made mandatory for each generator to be provided with capability to allow up to 105 % MCR( maximum continuous rating capacity) generation whenever situation demands. This is called Free Governor Mode of operation (FGMO).


Turbine Speed Droop
Here is the calculation to find speed droop.

s = [(No - Npr)/Nr ]x 100%

where
s = speed droop
Nr = rated speed
Npr = speed at full load
No = speed at no load

for example
Nr = 3000 RPM
Npr = 3000 RPM
No = 3150 RPM
s = 5%



Please continue read to http://anto-hendarto.blogspot.com/2011/12/load-shedding.html

No comments:

Post a Comment