kW Demand-based Load Shedding Scheme
kW Demand-based Load Shedding Scheme
(OP)
In the plant design, we have on-line Gas Turbine generator and a Steam Turbine generator. These two on-line in-plant generation facilities are separate plant locations. They are in parallel with utility transformers and are only designed to operate on plant load and do not intend to export power to utility. These on-line generators are there toi support plant operations in case, one or two of the 3 utility transformers are out.
We do intend to have a load shedding scheme in place using under-frequency relays (device 81). However, I also to intend to propose a "power demand" based load shedding scheme. That is using on-line kW demand on each power meters on each generator sets, I will be able to set a 1st stage alarm based on kW demand and then a 2nd stage to initiate load shedding. The under-frequency load shed initiate will be in an OR logic with the 2nd stage generator power demand.
I would appreciate if anyone can share their experience to this set-up or direct me to a link where I can get literature about this set-up.
We do intend to have a load shedding scheme in place using under-frequency relays (device 81). However, I also to intend to propose a "power demand" based load shedding scheme. That is using on-line kW demand on each power meters on each generator sets, I will be able to set a 1st stage alarm based on kW demand and then a 2nd stage to initiate load shedding. The under-frequency load shed initiate will be in an OR logic with the 2nd stage generator power demand.
I would appreciate if anyone can share their experience to this set-up or direct me to a link where I can get literature about this set-up.






RE: kW Demand-based Load Shedding Scheme
I understand that under-frequency protection is used to protect the utility when the total load on the local utility or portion thereof forces a disconnect from the main grid. If nothing is done, the grid system will collapse and your plant will lose power. By forcing large industrial loads to drop off the local grid has a chance to recover. This is not a parameter to base load shedding on.
Consider monitoring the power drawn from the grid and using that signal to control the prime movers on the generators so that you don't inadvertently export power.
If it is more expensive to generate your own power, still use the grid KW demand to control the prime movers.
If generator power is cheap, KW signal operates to use maximum power from generators.
If generator power is expensive, KW signal operates to use minimum power from generators.
Bill
--------------------
"Why not the best?"
Jimmy Carter
RE: kW Demand-based Load Shedding Scheme
Going back to my OP question, the load shedding scheme we intend to have is to protect our in-plant GTG and STGs in the event that the utility supply lines are tripped. If the incoming supply lines are suddenly tripped, then the GTG and STG will get overloaded and the system frequency decreases. We have to shed some loads to protect both load and the generators.
RE: kW Demand-based Load Shedding Scheme
I have had some less than optimum performance from a DCS system, but in all fairness, that was a long time ago.
If you use some method other than frequency to detect loss of grid power you may gain several cycles in response time.
Bill
--------------------
"Why not the best?"
Jimmy Carter
RE: kW Demand-based Load Shedding Scheme
As I understand, this increased in frequency will be detected by the individual generators' frequency and speed sensors and their governors will make necessary adjustment to be in synch with utility. I assume the there would be a time delay between the event of over-frequency and the governor finally adjusted the GTG and STG speeds ( also the MW output).
Is the frequency and MW output adjustment fast enough? if there is a delayed response, what will be the effect of this to the loads and to the utility?
I appreciate everyone's take on this.
RE: kW Demand-based Load Shedding Scheme
I am guessing that since you are dealing with a HRSG setup, this is a large load. The utility source thus is fairly strong as well? A downstream fault would be cleared much quicker by its relaying (and backup relaying) than any frequency effects you would see on the frequency. If you are operating in island mode, that is a different story.
How fast you machines are able to respond to load rejection are unique to the machine and that information should be available from the OEM.
One thing to consider is the utilization of alternate settings groups on your feeder and generator protection relaying (assuming it is a newer digital relay). In this case, you can easily detect the loss of 1 or 2 utility feeds (via breaker aux contact) and send that signal to the feeder and generator relays to initiate new settings. These can be "tighter" and selectively dump the non critical loads until the utility is back on line and can support the additional load.
Section 3 of the document below discusses load shedding.
ht
The second paper provides quite a bit of useful information as well.
https://d
RE: kW Demand-based Load Shedding Scheme
Also, if all other generating units connected to the grid are on free governor mode, there will be lesser swings, IMO.
If you want, please get a copy of MIL-HHDBK-1003-7 and see the discussions on Load Shedding. Just Google it.
RE: kW Demand-based Load Shedding Scheme
RE: kW Demand-based Load Shedding Scheme
Plain under-frequency protection is unlikely to operate quickly enough to prevent a prime mover stall, so it is necessary to detect the rapid slowing down of the machine before it actually reaches the under-frequency trip point and initiate disconnection from the system as rapidly as possible. There are a few techniques used for providing rapid disconnection from the system, including Rate-of-change-of-frequency (ROCOF) and vector shift. There are plenty hits on Google for both of these.
----------------------------------
If we learn from our mistakes I'm getting a great education!