Jul-2014

Avoiding compressor system downtime

Developments in anti-surge technology make it possible to maximise process efficiency and optimise compressor function

Amit Saxena
Dresser-Rand

Article Summary

The Compressed Air and Gas Institute (CAGI) estimates that industrial facilities in the US waste up to $3.2 billion every year on energy costs due to poorly designed and maintained compression systems. Compressors are the key drivers of many of these systems and damage to them can often result in costly maintenance expenditures, increased downtime and an overall drop-off in system efficiency. With energy prices on the rise, the need for new and innovative control technologies that can help cut operating expenditures and boost process efficiency are more critical than ever.

Understanding surge
Surging is one of the most common causes of compressor damage. Because it is highly dependent on a number of different variables, its occurrence is difficult to predict accurately. Even with anti-surge measures in place, many compressors have to be used rather conservatively in order to prevent damage to equipment. This results in wasted energy and decreased productivity, making process optimisation virtually impossible to achieve.

Surging occurs when insufficient flow into the compressor and/or an increasing pressure rise across the compressor causes a condition in which forward flow is unable to be sustained. This results in a temporary reversal of flow within the impeller, causing a decrease in the discharge pressure and/or an increase in the suction pressure.

This rise in suction pressure allows the compressor to 
re-establish forward flow but, when it resumes, the resulting pressure differential again reaches a point where the compressor becomes unstable, flow is reversed and this cycle is repeated. This continues until a change is made in the process and/or compressor conditions.

Surging can cause serious physical damage to pumps, fittings, valves, pipes, and other ancillary pieces of equipment. Rotor shifting caused by the surge cycle can also destroy thrust bearings and, in many cases, operating temperatures can exceed allowable limits, causing compressors to overheat. Because of this, it is always important to have effective anti-surge measures in place.

Surge can be prevented either by blow-off or recirculation of flow in order to keep the pressure differential across the compressor at a level in which reversal cannot occur. The moment at which either of these actions need to take place is determined by the controller, which is designed to predict the point at which surging is imminent (i.e., the surge line) by measuring a function of pressure rise versus flow.

The challenge, however, is being able to define accurately the surge line over a wide range of operating conditions. Because this is so difficult to do, engineers generally have to err on the side of caution and use compressors in a very conservative manner, resulting in decreased throughput and low operating efficiencies.

The key to maximising compressor efficiency is to determine the surge line with a high degree of accuracy. In doing so, the workable limits of the compressor can be clearly defined and unnecessary recirculation of flow can be kept to an absolute minimum.

Current surge technology
The technology that drives many of the anti-surge applications used today is based upon the premise that for any given rotational speed the compressor surge limit flow will correspond to fixed values of polytropic head and volumetric suction flow. This generally holds true for single-stage compressors, but many multi-stage compressors deviate from this theory. This methodology also produces surge control maps with coordinate systems that are only partially invariant to inlet gas molecular weight, temperature and compressibility.

Due to the volume ratio effect, which affects the polytropic head—suction flow relationship, the temperature and molecular weight of incoming gas can significantly change the point at which surge occurs in a multi-stage compressor. As a result, anti-surge algorithms that fail to produce surge control maps with coordinate systems that are completely invariant to changes in the properties of an incoming fluid are subject to a wide margin of error.

If an anti-surge controller that was unable to compensate for the difference of two gases with different molecular weights is used on a particular compressor, it would either not be able to prevent surge under all conditions, or it would produce a control line located so far to the right that the compressor would be highly inefficient when dealing with heavier gases.

Solving the surge problem
Dresser-Rand’s Universal Performance Curve Coordinate System is an anti-surge technology that makes it possible to maximise process efficiency and optimise compressor function by eliminating unnecessary fluid recirculation and/or blow-off.

The system automatically compensates for changes in molecular weight, temperature, pressure, compressibility, and rotor speed to produce a surge control map that is accurate across all possible scenarios.

Under normal operating conditions, proportional integral (PI) control is used to operate the compressor. The PI control loop is used to compare the control set point to the operating point of the compressor and provides an output to the surge valve to prevent flow from decreasing below the control line. Under these conditions, surge control action is initiated at the control line by opening the surge valve. This prevents a further shift of the operating point to the left towards the surge line.

In the case of rapid reductions in flow, such as process upsets, three additional controls are implemented to prevent surge from occurring. The first control is a back-up line, which is located between the control line and the surge line on the surge control map. In the event that the operating point reaches this line, Dresser-Rand’s Open Loop Step Logic quickly forces the surge valve open to increase forward flow through the compressor.

The second control takes effect if the operating point of the compressor reaches the back-up line a certain number of times within a specified period of time. When this occurs, the control set point is shifted to the right via Dresser-Rand’s Set Point Shift Logic. The flow set point continues to be shifted until the cause of instability can be corrected. This action establishes a larger margin of safety from the surge line.

The third control is a variable proportional gain action that takes effect when normal PI control response is unable to prevent flow from dropping below the control line during rapid system upsets.

To prevent surge under these circumstances, Dresser-Rand’s Floating Proportional Control Algorithm is initiated and surge valves are opened before the operating point reaches the control line. When the upset has been stabilised, normal PI control is resumed.