Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

islanding operation of six distributed steam turbine generators in a high voltage network 2

Status
Not open for further replies.

scientif

Electrical
Mar 18, 2013
16
0
0
MA
Hello, I hope everything is ok for you.
First, I wish to thank you for the time and consideration you gave to my post.
I am currently looking at the islanding operation of six distributed steam turbine generators in a 60kV high voltage network.
Each generator is a salient pole synchronous machine and has typically these power values: 37MW and 47MVA. the distance between the generators is in kilometers.
So here are my questions:
- are there any control strategies adapted for this situation, so as to operate all of these generators in island mode (disconnected from the grid utility, supplying local loads)?
- if there is an increase or decrease in load demand while working in island mode, how can the six generators share the load power demand?
- should droop or isochronous control be used for the generators during the island mode?
- are there any load shedding techniques adapted for this situation?
I am still struggling with this, so any help is really welcome; websites, company products, articles, theses, etc.
Thank you very much for your help.
 
Replies continue below

Recommended for you

- are there any control strategies adapted for this situation, so as to operate all of these generators in island mode (disconnected from the grid utility, supplying local loads)?
Do any of these machines have black start capability?
- if there is an increase or decrease in load demand while working in island mode, how can the six generators share the load power demand?
State of the art> A central load controller and fiber optic links between each site and the master controller.
Old school> Droop control. Isochronous controls don't work and play well with others. They are most suitable for single islanded sets.
Still old school, a central operator may contact the operator of each site and request a change of set point.
eg: With 3% droop, a set point of 61.8 Hz will result in 60 Hz at full load. If the load is trending at 50% at set point of 60.9 Hz will result in a frequency of 60 Hz with 3% droop.
You may want to drop some units off-line under low load conditions.
Old school is not so old school as you may believe. The operators and those responsible for the machinery at each site may be reluctant to yield control of their turbines to a remote signal.
An alternate control strategy may be to fix the throttles for a fixed output of all but one machine and designate the last machine as the swing machine. The swing machine will respond to load changes and may be run in isochronous mode(Exception to the general rule).
The machines may be arranged to alternate between 10% output and 90% output, on the command of the central load control authority. The load control authority may be a person or a computer. The issue of commands may be by way of a phone call or completely automated.
That's the broad strokes. You can run some models and explore these concepts. Any more detail and I may be writing the thesis, a violation of site rules.



Bill
--------------------
"Why not the best?"
Jimmy Carter
 
The above comments relate to plants under a common ownership.
For individually owned plants, economic and enforcement concerns are much more important than individual control schemes.
If one plant generates more than its share of power as determined by the Load Control Authority, what are the penalties and procedures?
If a plant declines or delays to accept an order to reduce power generation, what are the penalties and procedures?
See:
Example. Somewhere on the web is a list of enforcement orders. One operator paid a substantial penalty due to a 20 minute delay in complying with an order to reduce power generation.


Bill
--------------------
"Why not the best?"
Jimmy Carter
 
Dear waross, thank you for your response.

waross said:
State of the art> A central load controller and fiber optic links between each site and the master controller
I think that's an adequate solution for this situation. the process of load control and island operation has to be completely automated, with very few operator intervention. but i didn't understand what you mean by central load controller and master controller? is there any such system in the market?
waross said:
An alternate control strategy may be to fix the throttles for a fixed output of all but one machine and designate the last machine as the swing machine. The swing machine will respond to load changes and may be run in isochronous mode(Exception to the general rule)
This is a good solution, but the only problem is when passing from grid-connected to islanded mode, the deficit in power is bigger than any generator could handle alone. so if only one generator responds to load variation, its response may be so slow that the frequency could go below some limit, and protection device may then be triggered.
waross said:
The machines may be arranged to alternate between 10% output and 90% output
Do you mean all six machines or only the swing machine?
waross said:
The above comments relate to plants under a common ownership
yes in fact that's a site with four plants. 3 plants having one generator each, and one plant with 3 generators.

Thank you so much for your help.
 
Hi:
I am confused about your questions:"islanding operation " and "passing from grid-connected to islanded mode" are two different scenarios.
Scenario I :Always operate 7 units 4 plant stations system supplying load in islanding mode.
Scenario II: During system normal conditions, 7 units 4 stations interconnect with the Grid assuming at 60kV and after forced outages, somehow the 7 units 4 plant stations will be islanded from the grid and continue to supply the local load.

Which one is the scenario you are studying
 
Dear QBplanner, thank you for your answer.
I'm referring to Scenario II: During system normal conditions, 6 units 4 stations interconnect with the Grid assuming at 60kV and after forced outages, somehow the 6 units 4 plant stations will be islanded from the grid and continue to supply the local load.
Thanks again, and sorry for the confusion.
 
Then it would be difficult for Droop control with one unit maunally balance the power and load. You may need AGC (We called it in the large system) depending on the load and generation during the time the 7 units and local load splited from the grid. for steam turbine thermal generaors it would be difficult because you may have to shed some units to maintain teh F and V or quickly reduce the steam flow for some Gas turbine maybe. you will have to do an extensive study to figure out under various scenarios between load and online generators and you may want to talk to some companies who provide AGC and EMS system to get an idea how to make the system work.

In our system, we have a customer with 7~8 hydro units and a fairly constant load interconnecting to out grid.we have certain generator /load shedding schemes for the customer if they loss the interconnection to the grid depending on the prefault generation output , load and interconnection transfer.But they are hydro units so it is easy to shed and restart. I am not sure if the customer has a AGC because their 7~8 units are in the same plant.
 
Old school, totally manual.
5 diesel generators, total capacity 22.5 M
The operators checked and recorded all values every 15 minutes. Oil pressure, coolant temp.; Amps, Hz and Volts; KW for each online generator.
The operators would then tweak the governors to correct to 60 Hz. All governors in 3% droop.
When the load increased to a set level another generator would be started and placed online. The governor would then be adjusted for proper load sharing. Then the frequency would be adjusted to 60 Hz.
As the load dropped generators would be dropped off.

For a grid connection, droop control works. If at 100% output and the load is lost, the frequency will stabilize at 61.8 Hz or 51.5 Hz depending on where you are (base frequency plus 3%). If the output is less than 100% and/or there is some local load that stays connected when the grid connection is lost, the frequency excursion will be less than 3%.
Simple isochronous responds first in droop and then corrects the error. On loss of the grid connection, the islanded units may go to isochronous mode if desired.
As I think about your layout, I more and more tend towards an independent control for each site, rather than one overall control scheme. The reason is that given the spacing of your generating sites, there will be multiple points of failure where the connection to the grid of one site, the local load of one site or both may be lost.
The system will be simpler and more reliable if stand-alone control is implemented at each site.

Overall control may be implemented by a remote signal to adjust the frequency set points at each site. (And, as a result, the output.)

But remember that the most technically excellent and fuel efficient control scheme for a one owner installation may be unusable for a multi-owner installation.

A very simple example is the case where two generators are running at 40% load. If these are under one ownership then in most instances one would be shut down and one would be run at 80% output. Under separate ownership both may remain running at 40% output.

Can you post a one line showing the sites, local loads and grid connections?


Bill
--------------------
"Why not the best?"
Jimmy Carter
 
Bill:

I don't doubt about you case with 5 diesel each one rated at 5MVA on one site operated by one or two excellent operators.Operator can check them if they are on one site.

But, in my humble opinion diesel generators are slightly different from steam turbines generators and are relatively easy to operate. Steam turbine are more complicated than a diesel from the whole process pint of view, water process, boiler, turbine, generator and switching station, control room, station service, coal storage, water cooling. In the place where I came from, to operate a 50MVA unit steam turbine unit requires min. 20 people.

Governor can only catch the frequency but can not regulate it back to 60Hz for example, then you have to manually do it. Nowadays, it would be difficult to find such a experienced operator around. I missed the old schools time.
 
Thanks for sharing QB.
Our operators were working for minimum wage in the third world. We had some wrecks. Management philosophy was that we couldn't pay enough to get excellent operators so why try, just pay for repairs.
So of the issues were more humorous than expensive. An operator woke up the supervisor at 2:AM because a diesel wouldn't stop. He had forgotten to open the circuit breaker. I got a call to repair the reverse power relay.
One operator couldn't get his head around meter multipliers. When he came on shift, his kW records on the unit with a different meter multiplier took a jump relative to his current records. Rather than try to retrain him, we used the discrepancy as an informal check that the shift change happened on time.
The time that an operator went to wake his supervisor in the middle of the night because of a high temp alarm was a little expensive. By the time he got back, the old V 12 had stopped by itself.
Many of the small standby diesels run in droop and the slight frequency change with load change is not corrected. As I mentioned our operators sometimes adjusted the frequency every 15 minutes. All the customers on our very small grid new to not depend on a mains power electric clock for the correct time.

Bill
--------------------
"Why not the best?"
Jimmy Carter
 
Hi Bill: I like your story. good to know these. It would be more difficult to operate this small system. these operator are much better than the operators nowadays set in the control room with AC , computers and remote everything.
 
QBplanner, it may be some help to scientif if you have time to comment on some of the back end issues when a large steam turbine losses a large part of its load. How fast can the boiler reduce output to match the load? Black starts? Other issues?

Bill
--------------------
"Why not the best?"
Jimmy Carter
 
I work for a utility that is not interconnected to any major grids. We have a number of power plants spread out over many many hundreds of kilometres, and operate all units with approximately 2-3% droop.

A single operator monitors the system frequency from a remote location, and can adjust the loading on any unit over a SCADA network to maintain the frequency with limits. In addition, he also monitors the difference between "system time" and the time on a GPS clock and keeps this difference to within a few seconds by adjusting the system frequency slightly. Generally, very little interaction is required from the operator as load doesn't really change all that quickly, and there is no real issue with a small amount of deviation from 60 Hz. The operator will generally just pick one unit and use that to adjust the frequency with every other unit base-loaded.

In one location, which used to be islanded even from our island, we have installed a fully automated generation controller which will start and stop units and load them up in what it is programmed to view as optimal dispatch. All generation is at one location for this site, it becomes more complicated when you need to consider communications failures between sites.
 
the steam turbine can reduce output to a minimum level in minutes level 15 years ago. Nowadays the new design may be even faster.
Blackstart a 40MVA or 50MVA coal fired steam turbine generator could be a problem becuase you will need at least 7% to as high as 12% station service to start the units which turns to be average 4~5mva diesel units onsite with fuel storage.
We normally use steam units run base load and hydro units to regulate frequency. in this case, with 7 steam units in different locations and suddenly turning from grid supply to islanded operation and still maintaining various load supply is a bit challenge as I mentioned before.
With the 4 different locations, i would suggest they have to have communications in place between generation plants,load stations and central control room with EMS, plus AGC on some of the units in order to survive the above scenario. unless their base load is fairly constant ,then they can try to do some other economic ways to maintain the load generation balance.
 
Some of the "old school" methodology was just blind stupid luck. I inadvertantly tripped a unit on an islanded systemconsisting of two plants, multiple units at each plant, but only one unit running at each plant that winter evening. The operator at the other plant just happened to be up walking his rounds around the CR and had walked over to the tie panel. Hearing his unit (steam turbine) try to take on all the system load (about double its capacity) he just reached out and grabbed the tie breaker switch and cut us loose. About that time one of my mechanics who was in the second plant on an errand for me to pick up a tool happened at that moment the turbine groaned to be walking along the firing aisle (oil fired unit) on his way to the shop and he, having worked his way up from fireman up to mechanic knew immediately what to do (the fireman was nowhere to be found) and immediately began to grab oil guns and slap them into burners. Blind stupid luck but the system didn't go black and we were able to get our unit restarted and back on line within a short amount of time (combustion turbine, combined cycle.)

I wasn't so lucky the next time that I tripped that same CCGT... we did have to put our black start unit to the test (and it failed initally), but that's another story.

rmw

PS: all the islanded plants I can remember in my early days had two clocks on the operator's panel, one on the system, and one battery operated (sorry, no GPS clocks in those days.)
 
Status
Not open for further replies.
Back
Top