WO2014191051A1

WO2014191051A1 - Detecting surge in a compression system

Info

Publication number: WO2014191051A1
Application number: PCT/EP2013/061283
Authority: WO
Inventors: Andrea CORTINOVIS; Mehmet Mercangoez; Paolo Belli; Daniel Lewandowski; Michal Orkisz; James Ottewill
Original assignee: ABB Technology AG
Current assignee: ABB Technology AG
Priority date: 2013-05-31
Filing date: 2013-05-31
Publication date: 2014-12-04
Anticipated expiration: 2015-11-30

Abstract

A method for detecting surge in a compression system (10) comprises the steps of: receiving first and second data selected from at least two of electrical data (60), mechanical data (58) and process data (76) of the compression system (10); statistically evaluating the first and second data by detecting a change between past first and second data and actual first and second data, wherein a first and a second statistical value is generated; and detecting that the compression system (10) is in a surge condition based on merging the first and second statistical value.

Description

DESCRIPTION

Detecting surge in a compression system

FIELD OF THE INVENTION

The invention relates to a method and a monitoring system for detecting surge in a compression system, a method for configuring an anti-surge controller of a compression system and a compression system and a use of this method.

BACKGROUND OF THE INVENTION

Gas compression systems are widely used in pipeline systems and other oil and gas applications to transport process gas between different locations.

Dynamic compression systems, with, for example centrifugal compressors, have different constraints on their operating ranges. In the following, only the effect called surge is considered. This constraint is an unstable operating mode of a compression system that occurs when the developed compressor pressure drops below the network resistance that the compression system is connected to. Surge corresponds to fluctuation of flow and pressure up to the point of flow reversal with possible adverse effects such as overheating or mechanical damage. The avoidance of surge usually may be a safety requirement for a monitoring system of the compression system with a compressor protection system.

The efficiency of a compression system depends on the operating point of the compressor, more precisely on the mass flow and the pressure ratio (outlet pressure divided by suction pressure). The region with maximum efficiency is often close to the surge line, which is a line in the mass flow/pressure diagram of the compressor which indicates the surge region of the compressor. Therefore, it is preferred to operate near the surge region without crossing the surge line. However, the location of the surge line is typically not known exactly before the commissioning phase of a compressor and the corresponding anti-surge controller and generally anti-surge controller commissioning may involve testing of the compressor to establish the location of the surge line. It is however also a common situation that two different commissioning engineers may test the same compressor and come up with two different surge lines due to the difficulty and lack of objectivity in the judgment of the onset and the occurrence of surge from individual process measurements.

Therefore, a method that can facilitate the detection and characterization of surge events with certainty, repeatability and in a standardized way can be very useful to ensure that anti-surge controllers are commissioned optimally such that neither a high risk is taken by locating the surge line too close to the unstable region nor is energy wasted by setting the surge line too far away from the high efficiency zone.

Operating near the surge region without crossing the surge line requires demanding control systems and a constant monitoring of the risk of surge events during compressor operation. In most of the cases, the anti-surge controller may be able to prevent the surge condition from happening. However, large disturbances may still push the compressor operation into the unstable region and may cause surge due to uncertainties in the flow regime. Successful and accurate detection of the occurrence of surge may be very useful in assessing the performance of the anti-surge controller and in case the performance is not adequate, adjustments to the anti-surge controller may be made to increase the safety margins in an online fashion. The detection of surge events may be important to take countermeasures or to initiate an emergency shut-down of the compressor. The timing may be crucial for protection and anti-surge control.

In the state of the art, the tuning of the anti-surge control system is adjusted e.g. by moving the (anti-) surge control line further away from the actual surge line. This is typically done for a temporary duration to be more conservative and avoid possible repeating surge cycles.

A correct detection of surge may often be very challenging due to significant noise, delays and time constants of typical measurements used for this purpose which are mainly related to the gas compression process, i.e. pressures, flow, temperature. Especially at low speeds, the oscillations in the process variables may remain almost undetected if the surge cycles do not repeat.

Common approaches for surge detection monitor electrical variables like inverter current or other fluxes of the variable frequency drive using average values and thresholds, see for example US6354806 Bl and US4940391. If a signal exceeds the limits around this average, a surge event is detected and an alarm is produced for the operators but no automatic correction or adjustment to the anti-surge control system is made. Other approaches exist where the derivatives of the electrical signals are monitored for thresholding rather than the actual value. The main disadvantage of electrical data monitoring is the high noise contents of the signals and the indirect relation to surge through the shaft of the system.

In US4594051, the differential temperature is measured upstream of the compressor inlet. This temperature gradient can be an indicator for surge detection but only when a number of surge cycles have taken place.

DESCRIPTION OF THE INVENTION

It is an object of the invention to improve the control and the protection of a compression system.

This object is achieved by the subject-matter of the independent claims. Further exemplary embodiments are evident from the dependent claims and the following description.

An aspect of the invention relates to a method for detecting surge in a compression system. Usually, a compression system may comprise a compressor for compressing a medium like a gas that is connected to a network for transporting the medium. The compressor may be driven by an electrical drive, which may be controlled by a controller or control system. For example, the method may be used for air-, C0₂_ or Nitrogen- compressors and all other type of compressors. The controller may comprise a monitoring system for detecting surge events and an anti-surge controller for preventing the compression system to enter into a surge condition or at least to direct the compression system out of a surge condition, when surge has been detected.

According to an embodiment of the invention, the method comprises the step of: receiving first and second data selected from at least two of electrical data, mechanical data and process data of the compression system. The monitoring system of the compression system may monitor different parameters/variables of the compression system such as electrical variables (of the electrical drive), mechanical variables (of the electrical drive, the compressor or a shaft interconnecting the drive with the compressor) and/or process variables (such as a flow rate and/or a pressure of the fluid in the network). In particular, the monitoring system receives at least two different types of data, for example electrical and mechanical data, electrical and process data or mechanical and process data. It is also possible that the system receives three types of data, i.e. mechanical, electrical and process data as first, second and third data.

The data may be measured by sensors in the compression system or may be estimated from other data.

According to an embodiment of the invention, the method comprises the step of: statistically evaluating the first and second data by detecting a change between past first and second data and actual first and second data, wherein a first and a second statistical value is generated. After receiving and storing of the electrical data, the process data and/or mechanical data, which may be generated with varying sampling rates, the data of different types may be combined to increase the accuracy of surge detection. In particular, the method may be based on detecting changes between the history (past values) and the actual values of at least two types of data and/or changes in some statistics of a statistical variable based on at least two types of data. The statistical evaluation may be performed with the history of the data stored in a memory of the monitoring system.

A statistical evaluation of data may mean that the surge detection does not rely on a model of the compressor and/or compressor system and/or that the evaluated data are not compared to predefined values, for example to a predefined surge map.

According to an embodiment of the invention, the method comprises the step of: detecting that the compressor is in a surge condition based on merging the first and second statistical value. After receiving and evaluating the data, the statistical values are merged together to form a consistent decision on whether the operation is in a surge condition or not. For example, the statistical values may be weighted and/or compared.

The method may improve the surge detection in the whole operation range, because of the more accurate extraction of changing conditions. This may result in accurate surge characterization and therefore improved control and protection.

According to an embodiment of the invention, the first data is statistically evaluated to generate the first statistical value and the second data is statistically evaluated to generate the second statistical value independently from the first data. In other words, each of the first and second statistical values may be associated with only one type of data value.

According to an embodiment of the invention, the first and second data are evaluated together with a multivariate statistical method, for example the first statistical value (and optionally the second statistical value) may depend on the first and second data. It is further possible to combine different types of data for evaluation. This may be useful for determining a surge distance and for a derivative of the surge distance, which may be defined from combined signals using compressor characteristics in different directions on the compressor map. It might also be possible to come up with a probabilistic measure that surge will shortly occur or not.

According to an embodiment of the invention, the statistical evaluation is based on a principal component analysis using the T2 metric or kurtosis, a mean and variance tracking, and/or a windowing with thresholds.

According to an embodiment of the invention, the merging is based on a multi-rate Kalman filter and/or a voting system. The merging of the statistical values may be based on a multi-rate data fusion which can be achieved using a multi-rate Kalman Filter and using estimation error covariance as indication of abnormality, a voting system e.g. weighted majority and/or other indexes which describe how abnormal the operation of the compression system is.

According to an embodiment of the invention, the method further comprises the steps of: storing the first and second data in a data history; and/or storing the first and second statistical values in the data history. It may be possible that only actual data are received. To statistically evaluate the history of the data, the received data may be stored in a data history, at least for a specific time window or a variable time window.

According to an embodiment of the invention, the method comprises the step of: triggering further events, when surge has been detected. After the monitoring system has detected that the compressor entered into a surge condition, the monitoring system then may decide which further action should be triggered and may select them directly. Several interactions with further systems are possible, for example with:

(i) a commissioning tool for the anti-surge controller,

(ii) the anti-surge controller,

(iii) a protection system,

(iv) a process controller and/or a load sharing controller, and/or

(v) an alarm system.

The detected surge and the related statistical values (i.e. changes in the first and/or second data) may be used to trigger different actions like emergency shutdown, feed forward recycle valve control, and torque assisted anti-surge control. The interactions with other systems also may change control and/or protection settings such as:

(i) an anti-surge control configuration (for example setting a location of the actual surge line)

(ii) anti-surge control settings (for example shifting a location of control and trip lines)

(iii) process control and/or load sharing settings (for example shifting a location of a surge line constraint).

According to an embodiment of the invention, the method further comprises the step of: changing settings of an anti- surge controller of the compression system based on the detected surge. In particular, the configuration of an anti-surge controller may be changed automatically based on the surge detected by the monitoring system, e.g. by changing the location of the (anti-)surge control line.

According to an embodiment of the invention, the location of a surge control line, an anti-surge control line and/or a controller tuning such as controller gain or reset time are adjusted.

The location/position of the surge control line and/or the anti-surge control line may be altered/changed, when a surge event is detected outside of the surge region defined by the (anti-) surge control line and/or when the compression system is in a state which would imply surge as defined by the (anti-) surge control line, but no surge is detected in this state.

According to an embodiment of the invention, a direct emergency action is triggered based on the detected surge.

The different types of data that are received in the monitoring system may be electrical data, process data and/or mechanical data. It has to be understood that these data may be actual values, parameters and/or signals that are either measured by corresponding sensors or that are estimated from other data generated in the compression system.

For example, electrical data, which may comprise electrical data values, may comprise voltage values and/or current values of the electrical drive of the compression system such as a DC voltage, or a phase U current.

Mechanical data, which may comprise mechanical data values, may comprise position values, speed values, torque values and/or vibration values from a mechanical component of the compression system. Examples for mechanical data are a motor speed estimate, a motor torque estimate, a shaft power or a speed error in speed control loop.

Process data, which may comprise process data values, may comprise temperature values, flow rate values and/or pressure values of the compressed medium.

To summarize, possible surge events may be estimated and/or detected with strong certainty, repeatability and in a standardized way with the method by utilizing electrical data and signal processing techniques. The proposed method for surge detection may be coordinated with the protection functions and/or the anti-surge control strategy - for example by making active adjustments to the way an anti-surge controller (i) is configured (e.g. anti-surge controller commissioning) and/or (ii) is operated in real time (e.g. antisurge controller tuning).

A further aspect of the invention relates to a monitoring system adapted for detecting surge in a compression system, wherein the monitoring system is adapted for performing the method as described in the above and in the following. The monitoring system may be realized in a PLC unit running other control functions possibly including an anti-surge controller or on a stand-alone dedicated device that has the memory and computational means as well as connections for data exchange means. Selected results of the method may be displayed at an operator system showing for example the monitored metric and the degree of violation of a threshold or other rule along with the signals that had the highest contribution to the violation. Other computational logic may be made part of the monitoring system that scans for outliers or sensor failures to prevent unnecessary action on the anti-surge controller.

A further aspect of the invention relates to a compression system, which comprises an electrical drive mechanically connected to a compressor, which is adapted for compressing a medium and a monitoring system as described in the above and in the following.

The compression system may further comprise at least two different sensors selected from an electrical sensor, a mechanical sensor and a process monitoring sensor, which are used to generate the first and the second data. For example, an electrical sensor may generate electrical data, a mechanical sensor may generate mechanical data and a process sensor may generate process data.

A further aspect of the invention relates to a method for configuring an anti-surge controller of a compression system. It has to be understood that configuring may refer to the adjustment of configuration settings and/or configuration parameters of the controller. According to an embodiment of the invention, the method for configuring comprises the steps of: recording a trajectory of process variables; detecting a point with surge on the trajectory by performing the method for surge detection as described in the above and in the following, and adjusting a (anti-) surge control line stored in the anti-surge controller according to the detected surge.

After recording the trajectory of operating points (process variables), additionally surge- indicating signals of the electrical drive (e.g., drive power) are used to determine the surge onset. Once the surge onset is known, it is possible to reconstruct at which operating point the surge did actually occur on the process side. This combination may result in more accurate (anti-) surge control line identification for anti-surge control.

According to an embodiment of the invention, the method for configuring comprises the steps of: automatically testing the compressor at different speed intervals between maximum and minimum compressor operating speeds. For example, the method may be performed automatically during commissioning of the compression system. Surge may be detected and the complete (anti-) surge control line may be explored without excessive surge occurrence.

A further aspect of the invention relates to a use of the method as described in the above and in the following for standardized commissioning, re-commissioning and/or service activities. In particular, for commissioning, re-commissioning or service of anti-surge controllers for compressors with variable frequency electrical drives, the whole process of (anti-) surge control line testing may be automated.

It has to be understood that features of the methods as described in the above and in the following may also be features of the monitoring system and/or the anti- surge controller as described in the above and in the following and vice versa.

These and other aspects of the invention will be apparent from and elucidated with reference to the embodiments described hereinafter.

BRIEF DESCRIPTION OF THE DRAWINGS

The subject-matter of the invention will be explained in more detail in the following text with reference to exemplary embodiments which are illustrated in the attached drawings.

Fig. 1 schematically shows a compression system according to an embodiment of the invention. Fig. 2 schematically shows a controller for a compression system according to an embodiment of the invention.

Fig. 3 shows a flow diagram for a method for detecting surge according to an embodiment of the invention.

Fig. 4 shows a flow diagram for a method for configuring an anti-surge controller surge according to an embodiment of the invention.

In principle, identical parts are provided with the same reference symbols in the figures.

DETAILED DESCRIPTION OF EXEMPLARY EMBODIMENTS

Fig. 1 shows a compression system 10 that may be connected to a network for transporting a gas via a process gas inlet 12 and a process gas outlet 14. The compression system comprises a compressor 16 mechanically connected via a shaft 18 to an electrical variable speed drive 20.

The electrical drive 20 may comprise a transformer 22, an inverter 24 and an electrical motor 26. The torque of the electrical motor 26 is transferred by the shaft 18 to the compressor, which compresses the gas from a suction pipe 28 connected to the inlet 12 into a discharge pipe 30 connected to the outlet 14.

The process gas flows through the suction valve 32 in the suction pipe 28 to the inlet of the compressor 16. After being compressed, the gas leaves the compression system 10 through the non-return valve 34 and the discharge valve 36 in the discharge pipe 30. The non-return valve 17 is a protection device, which prevents back flow conditions at the outlet 14 of the compression system 10 and usually cannot be manipulated.

The suction valve 32 and the discharge valve 36 may be automatically manipulated by an actuator and therefore may comprise valve inputs 38, 40 for receiving control data. A cold recycle pipe 42 and a hot recycle pipe 44 introduce the possibility of cold and hot recycle. These actions are influenced by the cold recycle valve 46 and the hot recycle valve 48. An opening of the valves 46, 48 is automatically manipulated by the respective valve actuators with the valve inputs 52 and 50. In the cold recycle pipe 42, a heat exchanger 54 is needed to cool down the gas, which was heated by the compression.

The motor speed and torque may be adjusted by means of the drive input 56, for receiving control data for the drive 20. The electrical drive 20 provides mechanical signals or mechanical data 58 regarding compressor speed, drive torque, shaft power, motor speed and speed error as well as electrical signals or electrical data 60 like DC link voltage, phase U current and drive fluxes. These signals/data may be referred to as drive signals/drive data 62 and may be measured with electrical sensors of the drive 20 and/or may be estimated from the measured electrical data.

The shaft 18 connecting the electrical drive 20 to the compressor 16 is equipped with vibration and position sensors 64, which provide further mechanical data 58.

A temperature sensor 66, a flow sensor 70, and a pressure sensor 68 are connected to the suction pipe 28, whereas a pressure sensor 72 and a temperature sensor 74 are connected to the discharge pipe 30. All the sensors 66, 68, 70, 72, 74 provide process data 76.

Fig. 2 shows a controller or control system for the compression system 10, which comprises a monitoring system 82, a decision block 84 and a control block 86.

The monitoring system 82 receives the different types of data 58, 60, 76 from the drive 20 and the sensors 64, 66, 68, 70, 72, 74. After the data acquisition at different sampling rates, the monitoring system 82 processes the data 58, 60, 76 and performs an analysis, which is based on the comparison between a history in memory and the actual data 58, 60, 76.

Subsequently, a logic block/decision block 84 decides the actions that should be executed. Finally, in the last block 86, which comprises an anti-surge controller 88 and a protection system 90, control signals/control data are computed and sent to the actuators 56, 38, 40, 50, 52.

Additionally or alternatively, in the control block 86, configurations and/or settings may be adjusted.

Fig. 3 shows a flow diagram for a method for detecting surge in the compression system

10.

In steps 100a, 100b, 100c, electrical data 60, process data 76 and mechanical data 58 of the compression system 10 is received. The data acquisition is separated in three parts: In step 100a, electrical data is read in, in step 100b, process data is read in and in step 100c, mechanical data is read in. In steps 102a, 102b, 102c, the data is stored in a data history 104a, 104b, 104c. The data 60, 76, 58 may be separately stored into a respective electrical data history 104a, process data history 104b and mechanical data history 104c, which may be a window buffer storing the last N measurements and/or the last N statistics.

In steps 106a, 106b, 106c, the data 60, 78, 58 is statistically evaluated. A statistical value 108a, 108b, 108c is generated form the past data values stored in the data history 104a, 104b, 104c and the actually received data.

Each (or at least one of) the statistical values 108a, 108b, 108c may be generated by only evaluating the associated data 104a, 104b, 104c. However, as indicated by the dashed line, the analysis may be based on the other data 60, 78, 58 and data histories 104a, 104b, 104c, too.

Also the statistical values 108a, 108b, 108c may be stored in the respective data history 104a, 104b, 104c.

In steps 110a, 110b, 110c, a change in the data 60, 76, 58 and/or the respective statistical value 108a, 108b, 108c is determined. If such a change is bigger than a threshold value, this may indicate the occurrence of a surge condition. For example, after analyzing the current data point with the last N-1 history, the changes in statistics or properties are detected.

In step 112 the statistical values 108a, 108b and 108c (or the respective change values) are merged for detecting that the compression system 10 is in a surge condition. This step may be seen as a multi-rate signal fusion. In this step, the statistical values may be weighted and/or compared.

All steps 110a to 1 12 may be performed by the monitoring system 82, which in the end may generate a decision, whether the compression system 10 is in a surge condition or not. The result of the monitoring system 82 may be input into decision block 84, which in step 114 decides, which actions may be selected or triggered.

For example, in step 116, an emergency shutdown may be initiated or in step 118, a protection system 90 may be started.

It is also possible that in step 120 the results of the monitoring system 82 are input into an anti- surge controller 88, which uses the results for controlling the actuators and the drive of the compression system 10 such that it exits the surge condition. Furthermore, it is possible that in step 122 the results of the monitoring system 82 are used for changing control settings of a controller of the compression system 10.

Fig. 4 shows a method for configuring the anti-surge controller 88, in which settings of an anti-surge controller 88 are changed based on the detected surge and the results of the monitoring system 82. The method may be performed during commissioning of the compression system 10.

In step 130, a trajectory of process variables is recorded. For example, during commissioning, the compression system 10 is automatically or manually guided through different states by actively changing different settings of the drive and the actuators. In this way, the mass flow rate and the pressure may be changed until a surge condition is generated.

In step 132, the monitoring system 82 detects a point with surge on the trajectory by performing the method steps as described above. In particular, the detected point may indicate at which mass flow and pressure the compression system 10 is in a surge condition or is not in a surge condition.

In step 134, an (anti-) surge control line stored in the anti-surge controller 88 is adjusted/adapted according to the detected surge. The (anti-) surge control line may be a line or an arbitrary polynomial function in the mass flow/pressure diagram of the compression system 10 which indicates the stability margin of normal operation of the compression system 10. The position of this line may be changed by the results of the monitoring system.

The compression system 10 shown in Fig. 1 is one possible standard arrangement. There are many variations using for example only one recycle pipe/path 42, 44, a blow-off valve instead of recycle pipes 42, 44, additional or fewer measured signals, etc. The methods may be performed with any of these variations and are totally independent of how many signals/data are manipulated or measured. With the method, information from multiple signals/types of data may be fused and analyzed. Specific anti-surge control settings such as the establishment of the (anti-) surge control line during commissioning or the readjustment of anti-surge control parameters such as the location of the surge control and trip lines, feedback and feed forward controller gains, tuning or weights after the detection of a surge event during operation may be automatically manipulated with the output of the monitoring system. While the invention has been illustrated and described in detail in the drawings and foregoing description, such illustration and description are to be considered illustrative or exemplary and not restrictive; the invention is not limited to the disclosed embodiments. Other variations to the disclosed embodiments can be understood and effected by those skilled in the art and practising the claimed invention, from a study of the drawings, the disclosure, and the appended claims. In the claims, the word "comprising" does not exclude other elements or steps, and the indefinite article "a" or "an" does not exclude a plurality. A single processor or controller or other unit may fulfil the functions of several items recited in the claims. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage. Any reference signs in the claims should not be construed as limiting the scope.

Claims

1. A method for detecting surge in a compression system (10), the method comprising the steps of:

receiving first and second data selected from at least two of electrical data (60), mechanical data (58) and process data (76) of the compression system (10);

statistically evaluating the first and second data by detecting a change between past first and second data and actual first and second data, wherein a first and a second statistical value is generated;

detecting that the compression system (10) is in a surge condition based on merging the first and second statistical value.

2. The method of claim 1,

wherein the first data is statistically evaluated to generate the first statistical value and the second data is statistically evaluated to generate the second statistical value independently from the first data; or

wherein the first and second data are evaluated together with a multivariate statistical method.

3. The method of claim 1 or 2,

wherein the statistical evaluation is based on a principal component analysis, a mean and variance tracking, and/or a windowing with thresholds.

4. The method of one of the preceding claims,

wherein the merging is based on a multi-rate Kalman filter and/or a voting system.

5. The method of one of the preceding claims, further comprising the steps of:

storing the first and second data in a data history; and/or

storing the first and second statistical values in the data history.

6. The method of one of the preceding claims, further comprising the step of: changing settings of an anti-surge controller (88) of the compression system (10) based on the detected surge.

7. The method of claim 6, wherein the location of a surge control line, an anti-surge control line and/or a controller tuning such as controller gain or reset time are adjusted.

8. The method of one of the preceding claims,

wherein a direct emergency action is triggered based on the detected surge.

9. The method of one of the preceding claims,

wherein electrical data (60) comprises voltage data and/or current data from an electrical drive of the compression system; and/or

wherein mechanical data (58) comprises position data, speed data, torque data and/or vibration data from a mechanical component of the compression system; and/or wherein process data (76) comprises temperature data, flow rate data and/or pressure data of the compressed medium.

10. A monitoring system (82) adapted for detecting surge in a compression system (10), wherein the monitoring system (82) is adapted for performing the method of one of the claims 1 to 9.

11. A compression system (10), comprising:

an electrical drive (20) mechanically connected to a compressor (16), which is adapted for compressing a medium;

a monitoring system (82) according to claim 10.

12. The compression system (10) of claim 11, further comprising:

at least two different sensors selected from an electrical sensor, a mechanical sensor (64) and a process monitoring sensor (66, 68, 70, 72, 74).

13. A method for configuring an anti-surge controller (88) of a compression system, the method comprising the steps of:

recording a trajectory of process variables;

detecting a point with surge on the trajectory by performing the method according to one of the claims 1 to 9.

adjusting a surge control line stored in the anti- surge controller (88) according to the detected surge.

14. The method of claim 13, further comprising the steps of:

automatically testing the compressor at different speed intervals between maximum and minimum compressor operating speeds.

15. A use of the method of claim 13 or 14 for standardized commissioning, re- commissioning and/or service activities.