# synchro machines-el. generator theory

Hello people, I hope somebody here will explain me theory behind el. generator and turbine's behavior. Recently I visited one power plant where I saw how el.generator is
synhronized on a power grid. I saw that steam was passed to turbine using regul. valves and when turbine's rpm was 3000 (50 Hz) generator, operators synchronized generator with power grid and started to raise generators power (MW). What I don't understand is how this works exactly. Onece when generator is synchronized on the power grid, that sam grid force it to rotate with same speed 3000 rpms or what? Also I saw that same valves are used to raise load in MWs, so how is it possible that with same valve you speed up turbine (more steam== more speed) until 3000 rpm, and when gen. is synchronized opening that valve further speed remains same, but power starts to raise? Can someone explain me this or suggest a good link instead?
Thanks
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>

Once a generator is electrically tied to the grid, it will almost always rotate at the same frequency/speed as the grid (I won't go into the exceptions because hopefully they don't occur too often).
This is because if the generator starts to speed up, it starts to 'lead' the grid. The angle between where the generator 'is' in its rotation and where the grid is in its rotation will start to increase. This angle is sometimes called the 'torque angle'. As the angle increases, huge amounts of current and power begin to flow between the generator and grid. The simple explanation is that this power will flow from the 'leading' system towards the 'lagging' system in an effort to bring the two systems back together. They 'want' to stay locked together and anytime one starts to deviate from the other, power will flow between them to try and restore them together.
Now, when the turbine-generator is *not* connected to the grid, the regulating valves control the amount of steam flow so that the torque produced by the steam as it flows through the turbines exactly matches the friction in the unit and the speed of the turbine-generator is controlled. Opening the steam valves slightly will allow more steam to flow, and thus more torque is developed in the turbine. With more torque trying to turn the shaft, and the same friction as before trying to slow the shaft, there is a net positive torque on the shaft. This net difference in torque causes the shaft speed to rise.
Once the turbine-generator is tied to the grid, opening the steam valves still allows more steam flow. And more steam flow still develops more torque in the turbine. And *initially*, this extra torque overcomes all the other torques and the shaft *starts* to speed up. But as soon as it starts to speed up, now it is 'leading' the grid by a tiny amount. So the currents between the generator and grid rise, and more power flows from the generator to the grid. This power flow raises the amount of electrical 'drag' on the generator shaft (a torque applied in the opposite direction from the steam turbine's torque). This drag cancels out the higher turbine torque so the shaft stops accelerating. Result is, you have more steam flow, higher torque in turbine, but higher load on the generator and higher 'drag', and the shaft is still spinning at 3000 RPM, just at a slight 'leading' angle from where it was a moment ago.
In many large plants that are operated as 'base-load' units, the speed control signals for the steam valves are adjusted 'up out of the way' once the unit is tied to the grid. These controls no longer 'care' exactly what speed the unit is running at. A second signal, called 'load setting' is manually adjusted by operators to simply control the amount of steam flowing through the turbine, and therefore control the amount of electric power flowing from the generator to the grid.
Hope this helps...
daestrom
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Thank you so much daestrom, it's very good explanation and I understand that now. However, I have also one more question regarding active and reactive power that has been produced by generator. I know that, when there's more steam on turbine, more power is deliverd to the grid (and by that way, to consumers). By that power I assume active power and I think that power produced with help of steam is active. Since consumer can be more inductive or more capacitive in that case, in addition to active power, generator needs to produce reactive power. Since consumers are more inductive, there will be both active and reacitve power. In order to work, synchro machines (generator) needs to have "excitation" (I'm not sure if the term is correct since English is not my native language, and I'll appreciate if you correct me) that is DC current used to make magnetic field. I asked operators how generator produce reactive power and I got answer that reactive power is produced by using generator's "excitation". By changing DC current thay can control amount reactive power. Operators are people who operate machines based on exploitation manuals and usually, don't have any strong background in theory. I'm not really sure if this is true, and I don't understand how. Please can you explain how generator produce reactive power?
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>

Yes, that is 'active power' or sometimes called 'real power'. It is measured in watts (or kilowatts, or megawatts, etc...)

Reactive power is also sometimes known as 'imaginary power' after the branch of mathematics that uses 'imaginary' numbers (i.e. square root of -1).

Well, the operator got it basically right.
Capacitors and inductors conduct current that is out of phase with the voltage applied. These currents are called 'reactive' since they are due to capacitive reactance and inductive reactance. The amount of 'reactive power' is usually measured in VAR (or kVAR or MVAR). Reactive power is the rate at which energy is stored/returned in magnetic fields around inductors, or electric fields inside capacitors. Because it flows from the source (generator) to the load (inductor) and back again twice for every cycle of the applied voltage, no *net* energy stays transferred. It just 'bounces back and forth'.
'Apparent power' is the sum of 'active' and 'reactive', but because they are vectors that are displaced 90 degrees, they must be added vectorily. Because of the 90 degree displacement, Pythagorian theorem is useful....
Apparent Power = sqrt (real-power ^2 + reactive-power^2)
When one generator is supplying all the loads, you cannot control the amount of reactive load with the generator controls. Just like with one generator supplying all the loads, you cannot control the active load with the generator controls. The amount of connected loads and their type determine the total active and reactive load in this case.
But when a generator is connected to a grid where there are many other generators and many, many loads, it *is* possible to control how the loading is *shared* between all the generators. And that is what the operator was discussing. The number of generators and loads connected to a large grid is so huge, for most purposes it can be thought of as an 'infinite bus'. An 'infinite bus' will always be at a fixed frequency (one generator can't really change the grid frequency) and always at a fixed voltage (one generator can't really affect grid voltage very much).
Now, to understand how changing excitation will change how reactive load is shared among several generators, let me simplify it to a single-phase generator. The windings of a generator have an internal resistance, but also have an inductance. So we can analyze the machine by thinking of it as an ideal AC source with an inductor and resistor in series and the terminals where we connect to the generator are 'downstream' of the inductor and resistor. The amount of DC excitation determines the exact voltage of the 'ideal AC source' 'inside' the generator. But the terminal voltage is a function of that 'ideal AC source' *and* the voltage drop through the series resistance and series inductance.
In the case of a single, lone generator supplying a load, this internal impedance means that as load is added to the generator, the terminal voltage drops (more voltage drop across the internal resistance and inductance). So the excitation has to be increased, to raise the 'ideal AC source' voltage 'inside' the machine to make up for the larger voltage drop across the internal resistance and inductance when the load current increases. And that's one of the reasons why generators have 'voltage regulators'. To automatically adjust the excitation as load current changes, to keep the voltage at the terminals constant.
When a generator is tied to a large grid (approximately an 'infinite bus'), we have a fixed voltage on the bus and an 'ideal AC source' connected through an inductance. If we raise the voltage of the 'ideal AC source' (increase excitation), we have more current flowing through the inductance. Because it's an inductor, the current increase is not in phase with the bus voltage. So the generator is now operating with the current slightly out of phase with the terminal voltage. When the internal 'ideal AC source' is higher than the bus voltage, then the generator is supplying inductive current to the bus (it has a 'positive' reactive load). If the internal voltage is lower than the bus, then the bus is supplying inductive current to the generator (the generator has a 'negative' reactive load).
You may wonder where the 'real' load current comes into this. And you'd be right to ask. When there is just 'real current' (active loads only) on the bus, there is still a voltage drop across the machines internal inductance. So even when there is just active loads, the excitation must be increased a bit in order to compensate for the voltage drop across the internal inductance. So for the generator to have an active load and still have a 'positive' reactive load, the internal AC voltage must be even a bit higher than we thought before. If the internal AC voltage is *exactly* the same as the bus while the generator has an active load, because there is still a voltage drop across the inductance, the machine would actually have a 'negative' reactive load. (the grid is actually supplying the inductance inside the generator).
In large machines, the resistance is several orders of magnitude *smaller* than the inductance, so we can ignore the resistance without too much error. So we can simplify things a bit further as just an 'ideal AC source' and an inductance in series. The results are pretty much the same.
The final affect is what the operator told you. To increase the reactive loading on a single generator connected to a large network of machines, you simply increase the DC excitation. That raises the 'ideal AC source' voltage inside the machine, and more reactive current flows through the machine's inductance to the grid. The machine supplies more of the reactive load (and other generators located other places supply less).
There are some other things to consider. The electric 'drag' that a machine develops when supplying an active load is a function of the torque angle (as I mentioned in my last post) and the overall magnetic field strength of the rotor winding (the excitation). If you weaken the field strength, the torque angle must increase to maintain the same 'drag' on the shaft. But the variation of drag is actually proportional to the sine of the angle. So as the angle approaches 90 degrees, the amount of electric 'drag' on the shaft reaches a maximum. If the torque supplied by the turbine exceeds that amount, then the shaft *does* accelerate beyond the grid frequency. As the torque angle progresses from 90 to 180 degrees, the electric drag actually *drops*. From 180 to 360 degrees, the electric 'drag' actually works in reverse and accelerates the shaft. Then as the rotor goes around from 0 to 90 degrees again, the drag tries to slow the shaft again. This rapidly continues and degrades. It is called 'slipping poles' (the rotor poles are no longer 'locked' in with the magnetic field on the generator windings) and is a 'bad thing'. Sometimes it is also referred to as a loss of synchronism. To help avoid this, large machine voltage regulators have a separate circuit that will override the excitation and not let it be lowered too far depending on the *active* load that is being supplied.
But the real world is not an 'infinite bus'. If the generating station is located at the end of a fairly long transmission line (say, 50 miles), then the inductance and capacitance of the transmission line become an issue. If another large generator suddenly 'trips' off, the voltage on the transmission line may suddenly change. Or, if the line is actually two lines in parallel and one of the two lines has a momentary fault (such as lightning strike), one line may 'trip' while the other remains. Now the inductance, capacitance, and power carrying capability of the line has changed, and the voltage at the generator may suddenly change.
These sorts of things not only affect the voltage delivered to customers, but they also affect the amount of reactive power flowing through that internal inductance. And that can change the torque angle needed to provide enough electric 'drag' on the shaft. So a sudden voltage rise at the generator terminals (perhaps 1/2 the line tripped) could require the torque angle to suddenly increase and if it goes beyond 90, 'bad things' happen.
Grid systems are analyzed for likely scenarios and operators are given direction regarding how much reactive load to carry for a given amount of active load in order to avoid slipping poles and other 'bad things'.
Here in the northeast US, on August 14 2003, there were several lines that tripped. But the severity of the losses caused several stations to lose synchronism (i.e. slip poles). When they tripped, the transient got worse and more and more stations became unstable. So many transmission lines tripped that some parts of the 'grid' became so isolated that they became simple 'islands' with just one or two generators supplying some local loads and were not part of any 'infinite bus' anymore. Some parts of the grid that were further away from the transient (such as the state of Pennsylvania) detected the unstable situation and disconnected from the affected area and did not lose their generators. Others, such as New York City tried to 'hang in there' and ride out the transients and as a result their generators all tripped.
The 'internal inductance' I've been talking about is actually a much more complicated issue. The shape and type of rotor iron can have an affect on it. The current in the stationary windings creates a magnetic field that interacts with the field on the rotor, changing the overall magnetic field (this is a *big* factor and is referred to as 'armature reaction' because the winding on the stationary part is called the 'armature' winding). Also the depth of penetration into the iron of the magnetic field. The overall affect for 'slow' changes is lumped together into one inductive term known as the 'synchronous reactance' (for some rotors, there may be a 'direct' and 'quadrature' values, but not usually for steam turbine machinery). For rapid changes that involve magnetic field chanages that don't 'penetrate' all the way through the iron there is 'transient reactance' and 'sub-transient reactances'. But for a 'first year' class on AC machine theory, a simple 'series inductance' is enough :-)
daestrom
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
What a lucid explanation Daestrom. A good job indeed! Gene
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
I concur
--

Don Kelly snipped-for-privacy@shawcross.ca
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>

Well, then I know I must have got it right :-)
daestrom
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
I don't know if you're a teacher, but your explanations are great and one can learn stuff that is not in the books. I think it's bad thing not to share your knowleadge and experience with others. Thans again.
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>

Well, did that for a while :-)

But that's exactly what I'm doing, sharing :-)