The ultimate goal of compressor performance control is to match the compressor delivery to the process demand. This is achieved by varying the compressor throughput, which is why the term “performance control” may be also be called “capacity control”.
This article is an extract from the paper, "Compressor loadsharing control and surge detection techniques" by Wayne Jacobson, Medhat Zaghloul, Michael Tolmatsky and Serge Staroselsky of Compressor Controls Corporation, and Jeff McWhirter, an independent consultant at the 2016 Turbomachinery & Pump Symposia.
There are two aspects of compressor performance control that must be considered: The selection of the most suitable process variable to represent the process performance demand side of the equation; and, the selection of the most suitable mechanism to vary the compressor delivery side of the equation.
Some processes have wildly variable demand in a given period, while the compressors supplying that process have an unlimited, or almost unlimited supply sources. Typical of such a system would be a compressed air distribution network in a plant. In such cases the most suitable process variable is the pressure at the compressed air delivery manifold, which is the discharge pressure of the compressor(s).
To smooth out any sudden and large variations in the discharge pressure measurement reading, it is advisable to install a buffer tank (holding capacity) where the discharge pressure is measured. Another process/compressor configuration would be delivering compressed gas into an almost unlimited discharge volume, with a process demand that is “fixed” for long periods of time, but with the gas source experiencing fluctuations in supply.
As this is almost the opposite of the previously described one, it should come as no surprise that the most suitable process variable for good performance control of the compressor would, in this case, be suction pressure. A typical example would be gas collection from offshore or onshore gas fields and compressing it into a long pipeline.
Some processes require a certain supply of compressed gas for a specific reactive process, or to mix it in a certain proportion to another process fluid. An example would be compressed air in a blast furnace (steel plant) or air for a catalytic cracker (refinery). In cases such as these, then probably discharge mass flow of the compressor would be the most suitable process variable for compressor performance control.
The role of any centrifugal compressor used in a refrigeration cycle is to compress the refrigerant gas sufficiently so that it can be condensed with an economically available coolant, such as river or sea water, or using ambient air in air-cooled condensers. The ambient air or water temperature will dictate the pressure at which the compressor’s discharge gas will condense (or change state to a liquid refrigerant), so controlling discharge pressure would be inappropriate. Compressor suction pressure control would be a much more representative primary process control variable.
Control Element Selection for Performance Control Discharge Throttle Valve
Assume that it is required to control the pressure to a process using a discharge throttle valve on a single speed compressor. In order to accommodate process pressure set-point values that the process may require, it is necessary to force the compressor to operate at an even higher discharge pressure, and use the discharge throttle valve to drop the compressor discharge pressure to the value the process needs. So, even though the process feed requires only the expenditure of a compressor shaft power represented by P1, the compressor is required to operate with a shaft power equivalent to P2. This constant excess power consumed by the compressor makes this method of modulating compressor performance quite in-efficient.
Blow-off or Recycle Valve
Using the blow-off or recycle valve to control the process pressure. Here the compressor must operate all the time at close to its maximum flow capacity, with the excess to what is needed by the process recycled or blown-off.
Suction Throttle Valve
For single-speed compressors, it is a common practice to install a suction throttle valve that will, as it closes, lower the actual suction pressure of the compressor. If the discharge pressure remains relatively constant – as would be the case if the control intent was to control it – then as the suction throttle valve closes, the pressure ratio across the compressor rises. This, in turn causes the compressor to handle less flow.
On a compressor performance map using discharge pressure as the vertical coordinate, it is common practice to depict a “family” of performance curves for different suction throttle valve positions. It is important to keep in mind that a single speed compressor has only one performance curve – at the stated suction conditions, including pressure. As the mentioned earlier, when the suction throttle valve closes, an increasing pressure drop occurs across it, which means that the actual suction pressure of the compressor drops. However, it is convenient to depict the “family” of performance curves that represent the varying actual suction pressures on the original performance map at the “reference” suction pressure. Thus, as the suction throttle valve closes, it appears that the “reference” performance curve shifts to the left and downwards, for varying closing positions of the suction throttle valve.
Using a suction throttle valve results in the power consumption of the compressor varying with the varying process load, which makes this method of compressor capacity control much more energy-efficient that the previous two methods (discharge throttling and using the blow-off or recycle valve). However, there is slight energy penalty to be paid in the quantity of power lost through the inevitable pressure drop across the suction throttle valve when it is not in the 100% open position.
When selecting this method of controlling the performance of a compressor, the designer should be careful to ensure that the “reference” suction pressure – when the suction throttle valve is fully open – is sufficiently higher than atmospheric pressure, as any subsequent closure of the suction throttle valve would then produce sub-atmospheric pressures at the compressor immediate inlet flange.
Adjustable Inlet Guide Vanes (IGVs)
Another method to modulate the capacity (performance) of a single-speed compressor is to utilize adjustable or variable guide vanes located at the compressor inlet, upstream of the compressor’s first impeller. By modulating the IGVs, an angular deviation of the absolute velocity of the gas entering the leading edge of the first impeller is change, which causes a change in the head produced by the machine, the flow handled by it, and hence the power consumed by the machine. Using IGVs also results in the power consumption of the compressor varying with the varying process load. However, this method is even more energy efficient than the previous method (using an inlet throttle valve), because the energy losses associated with the suction throttle valve are eliminated. It is also possible to achieve a greater turndown of the capacity of the machine, while operating at high polytropic efficiency values. The additional energy savings thus obtained justify the higher capital and operating costs of the IGVs, which are more complex to build and maintain than a suction throttle valve.
Speed Variation
The most efficient method to modulate the capacity (performance) of a compressor is to vary its speed. Speed modulation makes full use of the Fan Law (for a constant diameter impeller), which state that flow is proportional to the speed of the machine, pressure or head is proportional to the square of the speed and power is proportional to the cube of the speed. Thus, the highest turndown is possible with speed variation, compared with other method of capacity control, and this allows for the most efficient energy reduction when the process load drops.
This method requires a much higher investment (capital costs) to provide the variable speed drive system for the compressor, and may involve higher maintenance costs than all other methods, but the energy savings over the life-cycle of the plant should easily justify this.