Three phase electrical power

[MBfreak]

Much of the electrical power we use today is generated by large prime movers ( turbines , mostly) driving large three phase generators.
However, an increasing amount comes from wind turbines ( each unit up to maybe 20MW) and large scale PV farms.
Germany is a prime example. Around 60 % of electrical energy per day often comes from windfarms and PV.

To start the technical part, here is a description of a large generator, the CB and step up transformer feeding the transmission network:

Hydro power plants usually run below 100 rpm and have generators with many poles. Some rotors are 15 m in diameter. The stator lamination is huge and transported to site in parts, assembled and most often the premade wirings are installed in the stator winding slots at site. ALL are wye connected. The neutral point is connected to ground thro a NER ( Neutral Earthing Resistor). A ground fault in the winding will be limited ta 15 A and NOT burn the lamination.
Voltages are 27,5 kV or lower and the stator winding is VERY complicated with several cooling systems built in. Winding are also transposed to get similar reactances in all parts. Theoretical and engineering knowledge levels are outstanding after > 100 years of manufacturing

Turbine poser plant generators usually run from 1500 to 3600 RPM, and power levels are 200 to 1900 MVA, ie 160 to 1600 MW. Most often hydrogen cooling and very complicated non-friction bearing assemlies. The design of the largest generators is half known and half guessed. The rotor can be 12 m long and 2,3 m in diameter at 1500 rpm. Rotor winding is a wonder, it has to stay equally distributed on the rotor at high speed and very large magnetic disturbances. I was involved in a plant that took the mfr 15 months to understand and rectify unbalances when rotor run at full power ( Excitation power 10 kV DC at 9,8 MW)
And that is easy compared to the very advanced math and physics needed to calculate and dampen the magnetic forces in the u-shaped winding transfer from one stator track to the next.
Generator is mounted on a 2 to 3 m thich concrete pad.

Large powerstation connect the stator phasor windings ( always below 27,5 kV) to a sometimes generator CB ( not common in the US) and then to a delta connected large power transformer ( often 3 off single phase units) thru aluminiom busducts where each phase has an outer earthed pipe, dia 1200 mm with a coaxial pipe conductor appr 600 mm. The magnetic forces in this system is impressive. Especially at a near short circuit fault in the HV transmission. Short circuit current may amount to 950 kA over many milliseconds. A 2 or 3 phase short in the busduct ( ie several earth faults) or in the LV side in te step up transformer will result in large scale damage where 1000 of kilos of aluminum is evaporated and generator end shields buckle. ( They are often 120 MM steel)

The generator is excited thru a large thyristor rectifier with VERY advanced control, to be compatible with network rules for stability control-
The generator has large current transformers and voltage transformers as input to a relay protection system with up to 50 different functions, from simple overcurrent to advanced computer based 100 % earth fault detection. ( Only available since 2009)
One very difficult issue is to keep a 1500 MW nuclear power station running for 250 ms at a dead short circuit very close to the step up transformers HV bushings.

Typical people involved in design, manufacturing, testing, installation and test runs:
Machine shops with super skilled people and machines
Plate shops with licensed people for hydrogen tanks with up to 100 cubic meter of hydrogen at 2,2 MPa
Winding manufacturer
Assembly workers
Heavy duty transport ( 600 tons, often by ship)
Riggers and installers
Test engineers
A few scientists that know everything about magnetic hi intensite bessel fields
Many design engineers, from how to run PT100 leads in hi magnetic fields to excitation rectifiers.

And why do i list all of this?

To try to make you respect some of the efforts that results in three phase power to your shop.

I am actually most impresse by the steam turbines. Last low pressure stage at 1500 rpm has peripheral speed og 2,3 time sound velocity.
Mechanical clearances very small. Impossible for me to understand


My knowledge on wind turbibes and PV farms is close to zero.

Next input will be a few snippets of theories and math for above

Ola

 

[cycle61]

Subscribed. And tagging in @slodat who also has years of experience in HV power.

 

[mike93lx]

Thanks Ola.

How do utility-scale turbines compare to the little ones that are in on-site cogeneration plants? My employer has a cogen plant running a Solar turbine spec'd at 3.5MW. We use it to make process steam and electricity.

Are they basically the same design, just scaled down?

We also have sites in NY with very old GE hydro turbines. They're fun to watch and I've been told, older than what GE has in their own museum.

 

[MBfreak]

Mike,
I do not have much knowledge of smaller plants.
However, a three phase 3,5 MW generator driven by some kind of turbine or combustion engine will in many respects be similar to the large machines.
The difference will be in the excitation system of the rotor winding.
The small ones , (if I am right?), often have inductive transfer from stator to rotor of the excitation power, which is then rectified by diodes in the rotor. This to avoid to have the sliprings and brushes , which wear and must be maintained.
BUT, it makes the excitation much slower, which is not possible in large machines.

Old hydro turbines are works of art.
There was an effort in the 90´s to increase efficiency of a large number of smaller (< 5 MW) old hydro power units in Sweden.
HOWEVER, measurements and calcs showed that they could only be improved very little. Effort was dumped.

Ola

 

[cycle61]

The inductive excitation is indeed common in smaller machines. This is a 3.5 megawatt engine driven generator, which we’ve partially disassembled the excitation wiring to perform some testing. DC is applied to a stationary set of windings at the end of the machine. An adjacent winding on the rotor picks up the field, which generates AC since it is spinning. Diodes then rectify this back into DC which is fed into the main rotor windings. It’s quite clever.



3.5 megawatts is a common size because at 480v it translates to approximately 4000 amps, which is the largest amperage readily available in low voltage breakers and switchgear.

This particular machine is 12.47kv but it’s a modular design which could easily have hosted a 480v alternator instead.

 

[MBfreak]

Here is number two, some theory and math.

Three phase "electricity" is where three currents and voltages from the same source feeds the same load.
And why three phase? ( there is also sixphase but nt common)
The three phases are 120 deg apart and there is no neutral current since the arithmetric summa current/voltage at any given time is zero
For those who find this interesting, there are EXCELLENT textbooks and Wikis that goes far beyond my abilities.

Since we maybe understand some aspects of electricity, there is a very solid underpinning in mathematics on virtually any part.
A few , with rather simple maths will take you a long way.
Waveform:
Always as close to pure sine waves as can be achieved. One period is a positive half cycle and a negative halfcycle. Both current and voltage goes thru two zero-crossings in each period. 60 Hz period length is 16,7 ms, 50 Hz 20 ms.

Phase displacement.
AC circuits with only resistance load has no phase displacement
AC circuits with inductive load or capacitive load has phase displacement. Since this can be correctly shown in a right angle triangle,
the displacemet factor is called " cosinus phi" The current displacement in the cap or inductor is + or - 90 degrees to the resistive part.

In a three phase system there are 3 different "powers"
S = U*I*sqrt3. This is what the system must push thru.
P = U*I*sqrt3*cosfi. This is the actual work that the generating system must produce
Q= U*I*sqrt3*sinfi. This is the power that is going back and forth in the system, not loading the generating system.
The transmission joule losses are of course added by the "Q current" So it is often minimized by inserting capacitors or inductors to
reduce the " bouncing" nonwork current.

Difference between phase to phase an phase to neutral voltage
Cosinus theorem will give you that phase -phase = phase -neutral* sqrt3

For all of the above you come a long way with high school math.

One irritating question is distorsion of the three equal and nicely separated three phase vectors.
Here you need a bit more math, if interested.
The three vectors may be unequal in length
The three vectors may not be separated 120 degrees.
Theory for this was worked out in the 1920 with rather demanding math.
A nice guy, Clarke , worked it out in a system called symmetrical components around 1945.
This made it accesible to lots of people. Like me.
In short
A distorted three phase system can be shown to be
-one symmetrical system rotating anti clockwise , phase a, b , c
-one symmetrical system rotating anti clockwise phase a, c , b
-Three rather small zero currents that may land anywhere within the three phase system vectors .
This leads to much increaed losses in a three phase motor.
Today you can put in data in a free solver on internet and get perfect values.
Which I, and old geezer, is much against.
Since you then not get to understand what you are solving , incorrect answers will gladly be accepted.

Magnetic circuits.
This is a difficult subject and often not understood well.
You can actually store a lot of energy in a large transformer when using aDC source to get a current and
voltagee thru the winding to determin resistance. No problem. R = U/I
However, you store L*I*I/2 joules ( Ws).
And when you gladly remove the battery leads a dangerous high voltahe flash will errupt
thanks to Lentz law, -e= dfi/dt.
And another tweak.
Large power transformers use heavy duty mild steel boxes to house transformer and oil. And need to have a certain short circuit impedance to reduce overcurrents at shorts outside. Often specified as ek= 12,23 % +- 0,09 % . Established in acceptance tests and tweaked in facyory if ouside specs, by manipulating distance of windings to steel box. Go figure! I have never understood it.

Electrical fields
Also a difficult subject for most people. But definitively something to treat with care.

Influence on living bodies?
Direct contact with AC can lead to dangerous burns and death. The current thru the body also disturbs internal organs.
The subject of being close to magnetic (B) and electrical (E) fields and influence on the body is much discussed.
And seems to be going nowhere.
Here is an example from a large power station with single phase busducts running 38 kA at 25 kV. The manual from 1985 stated since the coaxial busduct has 100 % of phase current as return current in the outer tube there is NO external magnetic field.
A measurement 2006 with a perfect instrument 3 m from the busduct metering in x/y/z showed large fieldsof several T .

So much for sundry physics stories and a little bit of math.
To sum it up, wiki has excellent articles on all of the above and MUCH more.

Ola

 

[RPH]

Excellent write up. This part I understand. At the power plants that called me in were all ready tore apart when I arrived. My part was insuring the induction machines for brazing the lead onto the field windings. They were very touchy about foreign objects. You had to prove no objects on you or in your pockets. The only parts that I missed and really wanted see was the removal and installation of the rotor core. That was a beast to balance from one end. All of the components you mentioned could be found in the induction equipment that Norway providing us. Later came the French, German, and Spanish design units. All similar but each a challenge in their own right.
Keep explaining Ola, this good for you and the rest of us.

 

[mike93lx]

Mike,
I do not have much knowledge of smaller plants.
However, a three phase 3,5 MW generator driven by some kind of turbine or combustion engine will in many respects be similar to the large machines.
The difference will be in the excitation system of the rotor winding.
The small ones , (if I am right?), often have inductive transfer from stator to rotor of the excitation power, which is then rectified by diodes in the rotor. This to avoid to have the sliprings and brushes , which wear and must be maintained.
BUT, it makes the excitation much slower, which is not possible in large machines.

Old hydro turbines are works of art.
There was an effort in the 90´s to increase efficiency of a large number of smaller (< 5 MW) old hydro power units in Sweden.
HOWEVER, measurements and calcs showed that they could only be improved very little. Effort was dumped.

Ola

Thanks. I'll have to ask the engineer for some info.

I'll try to grab some pics of the hydros when I am next on site.

 

[slodat]

I was the lead protection and controls craftsman in the largest power plant in North America for nine years. 4500MW from six units. Like the OP mentioned there was a push in the late 90’s to uprate where possible. The three bigger units had new stators installed to achieve the updating. I was there for the entirety of the full tear down mechanical overhauls as well.

My team designed, installed, tested, commissioned, and placed new protection in place. As part of the overhaul preparation new static excitation was installed on each unit as well as modern double derivative PLC governors, and replacing the original oil filled cable with modern overhead lines.

The quantities and magnitudes of these units is really difficult to wrap your head around until you spend a lot of time with them. Rotors weigh nearly 2000 tons, for example. That’s the rotor only, not the entire rotating mass of the 1,000,000hp turbine. Penstocks are 40’ diameter, and the currents are over 30,000a at 15kV.

I have a lot of experience with the big big hydro. Happy to share what I learned and answer questions. I can see the powerhouse from my backyard.

 

[MBfreak]

Now to my last input, which only relates to the < 24 kV power disribution systems. And ONEpersonal opinion.
The mass transfer of electrical energy is done by transmission companies at voltage levels above 170 kV.
The systems are much influenced by the country topology and climate. In some countries transmission is owned and run by the state.
There are some better than others. Hydro Quebec and the UK NGC are among the absolutely best. In all respects.

Distribution companies are often local and have wildly different technologies, knowledge and reliability.
Not in France, where EDF owns, builds and run everything electric
Now to the technical points.

Distribution have many transformer interfaces in substations like 145 to 10 kV. The 145 system is owned and operated by the transmission company. Transformers in the 20 to 100 MVA range and 10 kv switchgear is owned by the distribtion company.

10 kV is often built in ring so that a number of three phase transformers in the 1-4 MVA 10/0,4 kV are fed from two sides. If the cable or OH ring is damaged, most transformers remain energized.
The 10 kV ring is often an underground cable system in densely populated areas.
Rural distribution is often rather primitive wooden pole power lines, which is a major problem due to climate and trees close to lines.
In densely populated areas transformers and 10 and 0,4 kV switchgear are often built in concrete kiosks that are desigend to not be TOO ugly.
Rural transformers are often three phase 500 kVA or smaller mounted on the wooden poles that also carry the power lines.
All of these transformers are 400 V yn with solid earth connection, ie n to earth

Each property is fed by a three phase 400 V incomer ( ie phase to phase 400 V, phase to neutral 230 V) and then split up
with a fuse group and kWH meter to each customer. Which is ALWAYS indoor and remotely read by seller.
In many countries even a flat has three phase power with the panel in the flat.
It has an incomer disconnector and GFCI. No fuses only MCBs
If you are a true GJer and have a garage , three phase 400 V is bliss.
Compressor, lift, ventilation and AC can then run with 3 phase asynchronuos motors, which is far better than various one phase motor designs with capacitors, centrifugal switches and very low start torque.
No electric motor is as cheap, reliable and low mtce as a three phase squirrel cage rotor asynchronuos motor.
I once rebuilt 4 off 0,5 kW 1937 motors for a FUTUR-liner project. 100 % hobby job.
Single phase 117 V repulsion motors. Had never even heard the name before.

A few words about the 10 kV distribution system.
The system is most often impedance earthed so that ground faults are limited to 15 A to reduce risks for fire.
The transformer is D connected on the 10 kV side. The earthing impedance is a ZN connected reactor with a resistor added to earth from the N to get 15 A maximum.
This has a peculiar effect on the 10 kV system when one phase is affected by a low-ohm short to earth.
The two unnaffected phases will get neutral point offset so that phase to earth voltage as well as phase to phase voltage will be 10 kV.
Must be shut down very fast to prevent damage to consumers.

Typical IEC designation of a distribution transformer and ZN reactor are :
Vector group Dyn11
Voltage rating 10,25/ 0,4 kV ( IEC rating for insulation 12 kV)
Power rating 2000 kVA
Reactor ZN
Zero seq, impedance 30 Ohms
Added resistor to get 15 A earth fault around 600 Ohms, 15 A for 10 seconds.


And that ends my write up. On the electrical parts.
And now my ONE opinion
Mostly done to, as polite as I can, list my opinion of the problems you US GJ-ers have with complicated single phase electric motors and falling down wooden poles with single phase power lines and transformer up on top.


Ola

 

[RPH]

Good one Ola. Might want to mention the reactor function and give us a physical size on the resistors. I know some are huge!

 

[MBfreak]

There are excellent articles on the web on ZN reactors.
The ZN makes a connection from three phases to ONE neutral and impedance depending on voltage
relations between the three phases. Low impedance only when asymmetrical.
The NER is about 200 kg, 1*0,4*0,8 m.
Best are made by Metal Deployee, France.
All stainless steel and all connections welded

Ola

 

[micromind]

Here in Nevada, USA I've connected both the power and controls of a dozen or so generators that fed power into the electrical grid.

Most were 30-75 MVA (30,000 - 75,000 KVA) and operated at 12.5KV to 13.8KV. Some smaller ones, 5 - 15 MVA were 4160 volts. Most were 3 phase Y connected with resistor grounding. There was a switchyard nearby with a transformer slightly smaller than the generator output, this is because there are 'parasite loads'. These are loads needed to operate the plant.

A typical power system here would be the switchyard transformers high voltage side would be 120KV, 230KV, 345KV and 500KV. though there is still some 34.5 and 60KV left. These are transmission lines. Then there are substations that go from the aforementioned voltages to either 7200/12470 or 14400/24940 volts. Downtown there are still a few substations that go to 2400/4160 volts.

From these distribution lines, there are either pole-mount of pad-mount transformers. The vast majority of houses are fed with 120/240 single phase, solidly grounded neutral. These transformers are usually from 10 to 100 KVA. The larger ones feed multiple houses and/or apartments.

Commercial and industrial are mostly 3 phase, the voltages are 120/208Y, 277/480Y and 120/240∆. The Y systems are solidly grounded and usually up to 4,000 amps but I've seen a few 6,000 amp ones. Almost every 277/480 system will have at least one transformer that is either single phase 480 to 120/240 or 480 3Ø ∆ to 120/208 3Ø Y. Some of these are pretty large, I've seen them up to 1,000KVA.

The ∆ systems can have one transformer center-tapped and will have 120 from 2 phases to neutral/ground and 208 from the 3rd phase. These are solidly grounded systems. Another type of ∆ system is corner grounded or grounded B. This system is 3Ø 3 wire with one phase solidly grounded. These are useful because a single ground fault doesn't cause the entire system to de-energize, it simply causes an alarm. Mines use this system a lot. The 3rd type of ∆ system is ungrounded. Very rare these days but common many years ago.

Another interesting thing about the ∆ systems is they can operate with 3 transformers (closed ∆) or only 2 (open ∆). A lot of people do not understand ∆ systems and believe that an open ∆ has unbalanced voltages. Some do mainly because the 2 120 phases are heavily loaded and the high leg has very little load. The actual reality is that the vast majority of medium and high voltage switchgear has potential transformers that are open ∆. These transformers are used to reduce medium and high voltages to lower voltages that can be used with metering and protection relays. The voltage ratio needs to be very precise, if an open ∆ inherently causes voltage imbalance, it wouldn't work for this.

 

[cycle61]

And when using open delta potential transformers with a grounded B phase, it’s critically important to program your meters and relays correctly or else they will display an unbalance when the system is actually fine.

Many clients some technicians get confused by this.

 

[MBfreak]

In my humble opinion , the most confused person is the designer of a 3phase transformer connected in D and one corner grounded.

Caveat
Som large power transformers connected Yy have a third winding.
Delta winding to control 3rd harmonics. No power used.
To make it withstand short circuits in the Yy systems it has to be rated at about 1/3 MVA of the Yy system transformer
It is for that application common to earth one corner thru a CT or a VT to detect insulation problems in the D winding

Ola

 

[TRWham]

In my humble opinion , the most confused person is the designer of a 3phase transformer connected in D and one corner grounded.
...

Ola

It makes sense if you only need 3 phase line-line voltage for loads but still need to ground the system. You could use a center tapped XFRMR and ground the tap, but why buy that if you don't need the neutral and can just use a delta-delta with one line grounded?

 

[MBfreak]

TRWham
The electrically correct way to ground a three phas delta winding that supplies power to a grid is thru a three phase ZN reactor.
Often with a resistor from N to earth. ( in the 300 to 1500 Ohm range)
Fault current at a one phase direct earth on the three phase system thus limited to 15 A to avoid unlucky and sometimes devatsting fires.
An earth fault monitoring protective terminal shuts the transformer down within 15 secs.
Please note that the two "healthy " phases rise in phase-earth voltage to phase to phase voltage for the duration
And then :
In some parts of the world ( ie Sweden with very large flat granite surfaces where nothing grows or deserts with a totally dry sand surface) a phase wire at 4 to 24 kV may not result in a trip level of the earth fault terminal .
Special terminals must be used, often combining several indicators and they are reltively reliable.

Dir

 

[jar944]

Each property is fed by a three phase 400 V incomer ( ie phase to phase 400 V, phase to neutral 230 V) and then split up
with a fuse group and kWH meter to each customer. Which is ALWAYS indoor and remotely read by seller.
In many countries even a flat has three phase power with the panel in the flat.
It has an incomer disconnector and GFCI. No fuses only MCBs
If you are a true GJer and have a garage , three phase 400 V is bliss.

And that ends my write up. On the electrical parts.
And now my ONE opinion
Mostly done to, as polite as I can, list my opinion of the problems you US GJ-ers have with complicated single phase electric motors and falling down wooden poles with single phase power lines and transformer up on top.


Ola

Sigh.. if only.

And does anyone think single phase motors are a good/better idea? I always assumed they were a work around to the available power.

 

[micromind]

Sigh.. if only.

And does anyone think single phase motors are a good/better idea? I always assumed they were a work around to the available power.

3 phase motors are better in just about every way. They have only a frame, a stator, a rotor, some sort of cooling and bearings. Single phase motors have all that plus most of them have a start winding, usually a capacitor or 2 or 3..........and some way to disconnect the start winding. Far less reliable.