KO3 - Kompressor Optimization third generation
In 2005, NEA Group started the development process of a state-of-the-art reciprocating compressor design program. The goal was to have a program that covers all aspects of a reliable compressor design. The idea was that the program would be used within all departments and companies of NEA Group. That means that the sales, development, order handling, revamping and aftermarket engineers use the same tool for their specific technical tasks. The program should follow the compressor from its first inquiry to its final shutdown.
The program also should be able to cover all other brands and types of reciprocating compressors. This was a reasonable demand for Neuman & Esser since the aftermarket organization NEAC has the original equipment manufacturer (OEM) rights of various brands, which are no longer in the new machine business, but the compressors are still running in the field. Therefore, the program needs to have the flexibility to allow for designing all possible compressor configurations like horizontal opposed, vertical or V-type. In addition, the program should cover all possible stage arrangements. That means piston configurations up to eight stages on one throw should be possible for design, and even throw combinations with crossheads between cylinders were demanded. Multiservice compressors, as well as side streams, fixed intermediate pressure levels and tail rod designs were part of the specifications. Of course, the real gas behavior of all possible gases should be correctly predicted.
All these goals and more could be achieved!
Thermodynamics and compressor calculation
The compressor design always starts with the input gas analysis, the demanded design volume flow and the suction and discharge design pressure conditions. On the basis of this information, the program needs to be able to offer the ideal frame and stage configuration on the basis of the manufacturer’s frames and applied specifications like API 618 . The optimal frame and stage configuration is one that is highly loaded in terms of rod and crankshaft load and has good thermodynamic and mechanical efficiency.
Here, the appropriate valve choice is one of the key factors. Therefore, the program contains a detailed database of available valves that are an intrinsic part of the automated compressor design. The valve losses, valve clearance, as well as reliable valve lift are part of this design process, which takes place in the program.
After having the perfect compressor for the given design condition, all other design cases that need to be covered by the compressor must be checked under the same aspects. An overloading of the frame, especially, must reliably be excluded. Typically, varying suction or discharge pressure conditions as well as controlling the flow is demanded in the specification. KO3 covers all known control options like permanent or synchronized suction valve unloading, all possible clearance control, and speed control.
Verification of compressor valves
The various process conditions have an impact on compressor valve behavior. After having chosen the compressor layout, the valve design needs to be confirmed in terms of valve dynamics. For lubricated service, oil-sticking effects can have a negative impact on the valve dynamics. The intrinsic valve dynamics tool of KO3 also covers this aspect. Figure 1 shows the result of such a calculation with the dramatic suction valve oil sticking effect. It is immediately visible that a late closing of the suction valve due to sticking at the valve guard takes place. This effect leads to a reduction of flow and to higher impact velocities. The effect can be remedied by different measures like, e.g., stronger spring design.
For reliable compressor layout and verification, the inertia, gas and combined loads (rod loads) have to be considered. All kinds of loads have to be verified to be within the allowable load limits. The limits for all compressor parts are determined by detailed examinations including finite element analysis (FEA) and dynamic load tests. These limits are deposited in the program. To give a direct visual impression of the loading for a given rod and thermodynamic load case, a chart per rod is created.
The most sensitive compressor construction elements are listed in the chart for each rod shown in Figure 2. The user can easily verify if the compressor is well designed for the three load scenarios: design load, load at safety valve pressure and idle run. A possible overload of a construction element can immediately be identified and, if so, remedied by a design adjustment.
The yellow bars represent the load utilization of the individual components in terms of the loading at idle run, meaning the compression chambers are just loaded by the suction pressure of the stage. This is more or less the inertia load acting on the components. Typically, this load is not critical for low- to mean-speed compressors.
The blue bars represent the load utilization of the individual components in terms of the loading at design condition. The compressor parts are loaded by the gas load due to design suction and discharge pressure and by the inertia load. Of course, only the inertia load is taken into account, which really acts at the individual construction element, e.g., the crosshead is loaded by the inertia load of the piston, piston rod and crosshead.
The red bars represent the load utilization of the individual components in terms of the loading at the pressure safety valve (PSV) condition on the discharge side.
Details verification of compressor bearings by elastohydrodynamic
The different situations for crosshead pin bearings at full load and part load operation are another aspect. Sufficient rod reversal availability is always required for a highly loaded crosshead pin bearing.
Typically, the crosshead bearing is one of the most critical components in a reciprocating compressor. This bearing can fail as a result of hydrodynamic oil pressure being too high, the oil film thickness being too small at minimum or the rod load reversal being too little. The design program accommodates this fact by checking these three failure scenarios individually. If one of a crosshead, main or crankpin bearing seems to be slightly overloaded according to the chart, a detailed verification of compressor bearings by elastohydrodynamic (EHD) is immediately possible with the results is also very simple since one just has to compare the maximum hydrodynamic pressure and the minimum hydrodynamic film thickness with the critical reference values.
Figure 3 shows a cross-sectional view of the crosshead bearing’s EHD load situation under maximum connecting rod tension load. Both the small end eye bearing and the banjo crosshead pin are strongly deformed under the hydrodynamic pressure.
The deformation is so strong that the real hydrodynamic pressure needs to be applied for its calculation.
A simple representation of the rod loading by a single load would simplify the calculation process since the deformation would become independent from the hydrodynamic pressure. Unfortunately, this simple representation is not possible because it would yield such a dramatic deformation that interference between bearing and pin would occur. Therefore, the mathematical calculation process becomes more complex since the resilience of the structure needs to be implicitly introduced in the hydrodynamic solver.
Figure 3 shows that the deformation reduces the hydrodynamic peak pressure compared to a rigid structure since a more equally distributed hydrodynamic pressure takes place along the loaded side of the bearing. Consequently, the minimum oil film thickness also becomes less critical.
Since sufficient rod-load reversal is the most important requirement for the assessment of the reliability of a crosshead pin bearing, the EHD tool accounts for the refilling with fresh oil at areas where evaporation of the oil takes place. The brown curve in Figure 3 visualizes areas filled with oil and areas where evaporation has taken place and has not yet refilled.
Crankshaft strength and torsional analysis
The crankshaft load is generally dominated by bending and/or torsion. By making use of analytical bar models yielding nominal stress levels, their quantification can be carried out. Prospective critical locations are the fillets at the crank webs. The evaluation of the stress concentration there is the decisive task in this context. Because of the permanently varying rod load during a crankshaft revolution, the maximum stress in these fillets changes its circumferential position and magnitude all the time.
This effect can be best quantified by utilizing FEA models. They are used to adjust the local stress situation of the FEA with the analytical bar model by means of stress concentration factors (SCFs). Figure 4 shows an example of such a model.
Once a sufficient amount of FEA simulations have been performed, the results can be used to identify and adjust analytical approaches that produce approximately the same results as the FEA. That way, the individual crankshaft strength for a given job can be verified most accurately and quickly in the compressor design program KO3 without the need for intensive FEA studies. Figure 5 shows a model of a two-throw horizontal compressor.
For reciprocating compressors, it is mandatory to run a torsional analysis to assess the dynamic amplification of the torsional load on all components within the drivetrain. These are the crankshaft itself, the coupling, and the motor or engine shaft. In addition, the current fluctuation for a motor-driven train needs to be reliably predicted to protect the electric power supply grid. Therefore, it is mandatory to include the electromagnetic behavior of the motor in the mechanical vibrational system. The air gap torque fluctuations are not allowed to be neglected since they can have a dramatic influence on the torsional vibration.
The torsional vibration know-how needs to be core know-how for a reciprocating compressor manufacturer since the vendor of the compressor is responsible for the compressor train design in accordance with API 618. Neuman & Esser included, therefore, the complete torsional vibration tool into the design program KO3. Figure 6 shows the result of a typical torsional analysis in terms of Eigen modes, natural frequencies and dynamic energies.
Acoustical pulsation study
To avoid unallowable high pulsation levels when designing a new installation or a revamp, a damper check should be performed to determine the necessary volume of the pulsation bottles.
This damper check as well as complete acoustical pulsation studies can be run within the design program KO3. Numerically, the pulsations are calculated by the method of characteristics. The pressure and velocity pulsations are calculated for all calculation nodes in the time domain. Figure 7 shows the acoustical model of a two-stage compressor with two double-acting first-stage cylinders and one double-acting second-stage cylinder. The first-stage pulsation bottles have an internal choke tube, whereas the second-stage bottles are pure volume bottles. A cooler is acoustically modeled between the stages, which accounts for the strong dependence of the speed of sound on the temperature.
The user can visualize pressure and velocity pulsations at different nodes in one view. The P-V chart under the influence of pulsations is a result of such a calculation since the compression chamber is acoustically included in the model by means of an adaptive grid.
Fully automated design process and cost estimation
As long as the compressor vendor designs tailor-made compressor parts like the piston, piston rod, cylinder, cylinder covers and crankshaft, they need to be individually designed in accordance with the thermodynamic design. The piston especially is a highly loaded part that has to be assessed in terms of fatigue strength. A strength verification of the piston with an independent FEA outside of the KO3 tool would contradict a fully automated design process. Therefore, NEA decided to implement a simplified finite element method for strength verification of the piston in line with a predefined piston design standard. That means the piston must follow these design rules to make use of the internal FEA in KO3. Having this procedure prepared, a fully automated design process can be accomplished within the design program. The program determines all geometrical information of the parts designed by this process in accordance with the design standards. Of course, it is essential that those standards are available and are completely and explicitly defined. The design process is not completely straightforward since there are dependencies between the weight and, therefore, rod load, fatigue design and length of the piston. To overcome this difficulty, some calculation loops need to be programmed to find the final piston and cylinder design.
Besides the programming to determine all geometrical information, parametric designs of these parts need to be prepared within the 3-D computer-aided design (CAD) system. Once the parametric parts are available in the CAD system, the numerical values of all the parametric geometry data just need to be transferred from the design program KO3 into the CAD system. The parts are completely designed afterward without any additional manual work by the CAD engineer.
This automated design process has three advantages:
Since the complete compressor-designed process is also used during the project phase for an order, it is self-evident to automatically conduct the cost estimation based on these design data. This is part of NEA’s automated design process.
Compressor design for strongly varying pressure conditions
Some compression applications demand a very flexible compressor design in terms of varying suction and discharge pressure. One of the advantages of a reciprocating compressor is the fact that it can adapt to varying suction and discharge pressure conditions without a significant influence on the efficiency. Nevertheless, these varying conditions have influence on the flow, power consumption and rod load. To assess the complete influence in a single diagram, NEA’s design program automatically creates a so-called panhandle diagram (Figure 8). The abscissa of this diagram represents the varying suction pressure and the ordinate represents the varying discharge pressure. Isolines of constant flow (black) and constant power (yellow) are plotted in the diagram. Since the whole displayed operation area does not need to be necessarily reliable concerning all aspects of the compressor design, an examination of all operation points in the diagram is demanded. The green area in the diagram represents the operation modes where a reliable operation in terms of all design aspects is possible. Operation points in the red areas are not allowed since at least one criterion is violated. To understand which criteria are violated, a more detailed illustration is possible. An increased reliable operation area can be created by a design adaption with the detailed knowledge of the reason for the design violation. To create this diagram, including the compressor design, check that all design validations in terms of rod load, crankshaft load and thermodynamic design are carried out in a calculation loop.
Verification of measured rod loads and P-V charts
To monitor a running compressor, the NEA Group has the ability to measure the cylinder internal pressure to create a P-V chart and the rod load or crankshaft torque by means of strain gauges and telemetry system. To assess the gathered data, a simple verification in comparison with the theoretical prediction is necessary. Since the complete theory is implemented in the design program, it was obvious to create an automatic process of importing the measured data and visualization in common diagrams. Figure 9 shows such an overlay of a measured P-V chart in comparison with the theoretical one determined by the design program. In addition to the pure visualization, the measured data can also be used for estimation of the real behavior of the compressor. For example, the program can easily be determined by the real flow and power consumption. The real acting rod loads can be assessed, too. The measured data can be used to visualize and assess the rod loading in terms of the chart (“Rod Loading” section). The EHD calculation (“Details verification of compressor bearings by EHD” section) as well as the crankshaft loading (“Crankshaft strength and torsional analysis” section) can also be carried out with the real measured pressure or rod load data. Therefore, the design program is an excellent resource for the analysis of an existing compressor.