Controlling modern, low emission gas turbines can be handled by standard open control systems.
Gas turbines (GTs) almost always require automatic control systems, whatever their size and application. When starting, the unit auxiliaries, such as lubrication, ventilation and fuel, have to be brought into service sequentially. GTs are initially driven by the starting device; fuel is injected and ignited and acceleration continues until the minimum speed is reached.
Up to this point, control actions are more or less the same for all non-aviation applications, certainly in the range 10-30MW – power generation (generator drive), mechanical drives (compressors/pumps) and marine propulsion. The control system also has to provide an automatic stop and cooling sequence.
Then, when on load, all applications require speed control, in most cases combined with cascade control for the main process parameter – say electrical power. Again, it’s also the responsibiliity of the controls to provide load limitation, to prevent excessive temperatures and speeds, and to provide automatic protection.
Some GTs have a dual-fuel capability, being able to burn gas or liquid fuel, or a combination. Here, the control system must additionally be able to switch automatically between fuel types or mixtures. A local operator interface for control and monitoring is then a necessity, and remote control is often required.
GT control challenges
For this range of duties most GT controls are based on dedicated, specialised systems. However, some users are moving to standard distributed control (DCS) platforms. ABB Stal’s GT35 (17MW) and GT10 (25MW) turbines are examples. On these, the controls are based on ABB’s Advant and Master. One process station (PS), the MasterPiece 200 (MP), takes care of all controls except fuel governing, which is implemented in a second, smaller Advant Controller 110 (AC).
The AC has custom I/O for the fast fuel control analogue signals. The two PSs are linked with a Master fieldbus and an Advant 510 OS provides the operator interface – linked to the MP via MasterBus 300.
There are, of course, pros and cons with standard DCSs versus specialGT controls. Advantages include:
* Low costs, since the platform development, maintenance of software, hardware and documentation and technical support are shared.
* Continuous development is automatically provided by the vendor.
* High-level, function-block languages are included. They are self-documenting and facilitate implementation of special features, and integration of extra functions (such as overall control of several units, or switchgear supervision).
* DCS systems are highly expandable, alowing movement to much larger applications – like combined-cycle plant – and extended functions, such as information management.
Looking at GT controls in a little detail, there are several interesting aspects. First, with the growing concern for more environment-friendly power plants, ABB has develop dry, low-emission (EV) burners for GT combustors. Reliability here is a factor – there are no moving parts, these being very vulnerable at 600-1000srC. On the GT10, the EV burner ensures low emissions for gas fuel only – water or steam injection is currently required for NOx reduction on liquid.
This dual-cone burner requires dual control valves running in sequence. The main flame is lean-burning, and is only stable at high loads. A supporting primary flame is therefore used – this is the only flame present during start-up and at low loads. The transition begins at around 60% load, the exact point being determined mainly by the combustion zone temperature.
The complexity of this sequential control process is highlighted by the need for the fuel flow to be reduced from full (100% main flame) to around 8% within just 0.3sec.
During this transition, it’s vital that primary fuel is over-enriched. If the GT is running at full-load and its breaker opens, it’s expected, after a short over-speed transient, to run at synchronising speed. As the GT cannot be re-ignited at speed, a minimum amount of fuel has to be secured during the speed transient.
Next, to keep CO emissions within limits, the combustion zone temperature has to be maintained. Low loads, providing a relatively high GT air flow, force temperatures down – so air has to be diverted past the zone via the combustion chamber bypass control valves and a compressor bleed valve. The bypass valves have to be controlled in load regions below full load where the zone temperature starts to fall. At lower loads, the compressor bleed valve is then opened and air recirculated, increasing zone temperature and limiting CO.
Compared with basic combustors, in which fuel flow, controlled by a single valve is the only means of control, control of the EV combustor is complex. Not only must parameters lke output be control-led, but also emissions must be handled by controlling fuel distribution in the burner, and the combustion zone temperature. This needs two control valves actuated by a.c. servo motors, plus six pneumatically actuated closed-loop control bypass valves. It also involves pneumatic compressor bleed valves.
Control of fuel transition, from gaseous to liquid fuel and vice versa, in dual-fuel GTs is another control area. Single-fuel control is via controllers and limiters on a minimum selector – ie the controller requiring the smallest fuel flow has control. An exception is the minimum fuel control which is connected over a maximum selector.
With dual-fuel control, each fuel has its own selector set-up. Some controllers and limiters are common – eg speed controllers and temperature limiters – whereas others are dedicated – eg minimum fuel control, maximum valve stroke control and fuel change-over.
During normal running, the power turbine speed controller is harnessed. Looking at transition from a single gas fuel to liquid, after the system is activated, the governor sets the minimum liquid fuel flow and the control valve is positioned accordingly. As the PI controller for speed senses the sum of the stroke of the control valves as a feedback, the gas fuel control valve is reduced.
When liquid fuel is selected, the fuel change-over control inserts a maximum stroke signal as a negative ramp to the gas fuel control. As the gas control valve is closed, the liquid control valve is forced, by the speed control feedback, to open to compensate. The fuel change-over is completed within 25sec.
Transfer from liquid to gas is faster as the start-up time of gas fuel is shorter. Such is the tightness of control that fuel change-over is achieved with a power transient of less than 4%.
* The Author is with ABB Stal.