Control.com - Nerds in Control

Hi there,
Apologies if my questions are quite basic as
I'm quite new to the topic.

First, I'm unfamiliar with the term hunting.
I'm currently modelling a CHP unit with a gas turbine prime mover. The modelling is based on that provided by W.I Rowen that consists of three control loops (speed, temperature (and IGV)and acceleration. The aim of the simulation is to investigate the dynamic behaviour of the gas turbine when subjected to different modes of operation.

The one i'm currently simulating is two machines operating in parallel interconnected to the grid, with one machine on droop control (loaded to 0.7pu) and the other on isochronous control (loaded to 0.9pu). A load of 0.45pu is then added causing an overload. I've been asked to investigate the hunting between the two. I've had a look through many text books and articles, but have not been able to find one that describes the phenomena well. I've done a test simulation, and found that although the machine on droop control reacts as expected and picks up the required load, the machine on isochronous mode fails to pick up any.

Should this happen? I hope you'll be able to answer my questions. Thanks

It would seem your first problem is your assumption that when connected to a grid with other units that one turbine control system should be in Isochronous speed control mode and the other should be in droop speed control mode. No turbine which is connected to a large ("infinite") grid in parallel with other units should be operated in Isoch mode; period. Unless that unit is very large and has some special form of frequency control or some special control function. If a unit which is connected to a large grid is being operated in parallel it will almost always "hunt", which means it will very quickly increase and decrease its load, almost wildly, and usually some protective relay (most likely reverse power) will open the generator breaker or open the generator breaker and trip the turbine.

Next, if one unit were operating at 70% of capacity and the other unit were are 90% of capacity and the load on the two suddenly increased by 45% of capacity, that would exceed the total capacity of the two units (if I'm presuming the two units are similarly rated and the 45% added load is equal to 45% of the rated capacity of one of the units). In other words, if the two units were both rated at 40 MW each, and one was running at 28 MW and the other was running at 36 MW and suddenly the load on the two increased by 18 MW, the total load on the two would increase to more than 80 MW (28 + 36 + 18 = 82 MW).

First of all I don't know how such a thing could happen to a CHP interconnected to a grid, forgetting the fact that when connected to a grid with other generators that both units should be in droop mode. Second, it's not possible for a unit to operate stably in Isoch mode when connected to a grid with other generators, unless the Isoch mode is really de-tuned or some special form of Isoch is being used.

So, I think some of your assumptions are wrong about how generators operate when connected to a grid in parallel with other generators.

Hunting of a control device is something that is characterized by unstable process control and oscillations of the process variable, such as load swings on a turbine generator, or pressure swings on a pressure regulating valve. Instead of being stable, the device is unable to maintain a stable process variable (such as load or pressure in these examples).

When the control system of the prime mover of a generator which is connected to a grid in parallel with other units is switched to Isoch control, it will try to control the frequency of the grid with which it is connected. And in so doing, its power output, and in the case of a combustion turbine, its fuel valve will oscillate, usually wildly. You haven't lived until you've been on a ship where someone accidentally switched the steam turbine control of a generator which was in droop to Isoch and was operating in parallel with another generator which was in Isoch. The wattmeters just oscillate wildly, one full positive, the other down to zero, and the lights just flicker and motors start speeding up and slowing down. It's just crazy.

If you want to see the frequency of even a very large grid in real time, log on to www.ucte.org. There is a little applet at the top of the page that displays the European grid frequency in real time. Log on a different times of the day and night, especially in the morning when people are waking up and in the evenings as people are going to bed. That's when you'll see the biggest deviations from 50.0 Hz for the longest periods of time, until the grid operators either increase generation (in the mornings) or decrease generation (in the evenings) as load goes up (in the mornings) or goes down (in the evenings). Some days are "bettter" than others.

It's very instructive. And can be somewhat misleading, too. As you watch it, it will rescale itself as the deviations get bigger so that they won't appear as large as they actually are. And, at other times, when it re-scales itself to something like 49,991 Hz to 50.009 Hz, the fluctuations appear much worse than they are (+/- 0,009 Hz, which ain't much, really, but when expanded to entire y-axis of the graph appear to be very large!). It's really fun to watch, again, especially in the mornings and the late evenings when the biggest load changes are occurring as people start and finish their days. It shows just how difficult it can be for grid operators to control grid frequency during such times.

What I wish we could see on a concurrent graph is generation/load; because for grid frequency to be stable the generation must exactly match the load. If generation exceeds load, grid frequency will rise; if generation falls below load grid frequency will drop. So, in effect what we are seeing is a graph of generation versus load, because the frequency of a AC grid is a function of generation versus load, but I would like to see the actual MWs for each.

On such large grids, there is probably not a single unit which is operating in Isoch mode to try to control grid frequency. Rather. operators must raise and lower the loads of many units operating in droop mode, usually through something known as AGC (Automatic Governor Control), or something similar. Or by instructing plants, usually combined cycle plants, to start up or shut down. Their job is quite difficult, especially when large generators trip or when large blocks of load or generation are separated from the grid.