Synthesis of work exchange networks for gas processing applications 5

Chapter 5 Conclusions and Recommendations CHAPTER 5 CONCLUSIONS AND RECOMMENDATIONS 5.1 Conclusions In this work, the optimal operation of the latest AP-XTM refrigeration system is presented at different process conditions, MR compositions, and weather fluctuations. Though it is not implemented in the existing AP-XTM system due to proprietary restriction, the model is able to mimic realistic operation and generate optimal operational parameters so that operators can utilize it to fix parameters optimally. Extensive study can be performed if the existing AP-XTM process is available in the public domain. As it is understood that compressors are major users of energy in process plants, WENS is studied to utilize available energy in the process. Due to high nonconvex and complex nature of this process, this preliminary study is carried out at single speed. It is found that WEN is able to utilize available energy in the process economically. However, further study needs to be performed to synthesize an economical WEN. 5.2 Recommendations As both of the models are simplified and several assumptions are made due to complex nature of constraints, the following improvements are recommended. 1. One of the first recommendations for future works would be to have access to certain proprietary AP-XTM operational data such that this model can be updated with actual process parameters. Only then, can it be verified and further improvements are made if discrepancies are found. 105    Chapter 5 Conclusions and Recommendations 2. One area of possible improvement in the model is compressor performance modeling. The current model assumes that the compressor efficiency is invariant across all operating conditions. This assumption is only valid if the compressors are operated within a narrow capacity range and near their design point where efficiency is relatively constant. Although no significant changes in compressor capacity and/or factors that affect compressor performance was observed throughout the case studies, it is inevitable that the accuracy of the model is compromised to some extent, though minor. Therefore, as an improvement and to better reflect the actual efficiencies of compressor systems, a performance based efficiency factor could be implemented into the model formulation. This, however, necessitates the availability of performance curves of compressor systems in used today such as General Electric’s BCL and MCL series used by Qatargas II42. 3. Besides this, including compressor stage arrangement with inter-stage cooler can be also a potential improvement of this model; however, it will make the model highly nonlinear and MINLP. In that case, solver improvement to handle such a complex model is prerequisite. 4. As this process includes turbines and compressors in different cycles, the idea of WENS should be included so that an economically energy integration is carried out. 5. Furthermore, it is irrefutable that MR composition has great impact on system performance. To achieve maximum process efficiency, an optimum or near optimum MR composition must be utilized. This model in its current form does 106    Chapter 5 Conclusions and Recommendations not have the capability to optimize the MR composition but allows the user to iteratively set different MR compositions in the quest for lower power requirements. An extension to the current system should include a model that automatically generates the MR composition for which there is the greatest match between the hot and cold composite curves. 6. The first recommendation for WENS is developing model with variable speed that will offer optimal synthesis of this network. This is not executed in this work as highly complex nature of the operating curve is the main barrier. 7. Due to unavailability of operating curves for different mover model in the public domain, fictitious operating curves are used to demonstrate case studies. Therefore, to compare with the real process, it is prerequisite to obtain the performance curves for the existing movers available in the market. Besides this, accurate cost parameters need to be obtained. 8. This model is developed based on a single operating curve. However, including all the available operating curves in the model will make it more realistic. 9. Only one SSTC is utilized in this model. However, multi SSTCs need to be considered so that movers are arranged economically in different SSTCs. It is worth to mention that arrangement of movers in SSTCs and their number in one SSTC increases vibration and thus noise level. Government rules and regulations are the main constraint of this placement. 10. Availability and spare philosophy are not considered in this model. These can be included in the model to be competitive with the traditional industrial practice. 107    Chapter 5 Conclusions and Recommendations 11. As the abovementioned improvements demand high computational work, it is imperative to improve existing commercial solvers so that this nonconvex problem can be optimally solved. 12. After solving the variable speed model, nonlinear operating curves (Pressure Ratio vs. Corrected Flow) can be utilized to synthesize a real WEN that will facilitate the study reducing the energy consumption of the process. 13. Finally, variable efficiency of corresponding equipment should be included to obtain realistic operation. 108     ... efficiencies of compressor systems, a performance based efficiency factor could be implemented into the model formulation This, however, necessitates the availability of performance curves of compressor... composition for which there is the greatest match between the hot and cold composite curves The first recommendation for WENS is developing model with variable speed that will offer optimal synthesis of. .. synthesis of this network This is not executed in this work as highly complex nature of the operating curve is the main barrier Due to unavailability of operating curves for different mover model