Considerations and complications with "first principles" dynamic modelling of industrial compressors
Readily available processing hardware and "off-the-shelf" (OTS) simulation software has made "high fidelity" first principles models of both steady and transient states, for both axial and centrifugal industrial compressors, relatively easy to construct. These high-fidelity model...
Guardado en:
| Autores principales: | , , |
|---|---|
| Formato: | article |
| Lenguaje: | EN RU UK |
| Publicado: |
National Aerospace University «Kharkiv Aviation Institute»
2021
|
| Materias: | |
| Acceso en línea: | https://doaj.org/article/e0799c6f3ec74c088b6f354e4139aba4 |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| Sumario: | Readily available processing hardware and "off-the-shelf" (OTS) simulation software has made "high fidelity" first principles models of both steady and transient states, for both axial and centrifugal industrial compressors, relatively easy to construct. These high-fidelity models are finding their way into "real-time. digital twin" performance monitors, front-end engineering design, and post-design – pre-construction compressor performance evaluation. The compressor models are useful for reliably demonstrating the compressor and – to some degree, based on the complexity of the model – process response to various operating conditions. Once the model is constructed, it is trivial to run a "what-if" analysis of compressor performance to answer questions related to (a) recommendations or validation of the recycle/vent valve size and actuation speed, (b) general piping layout and sizing around the compressor, (c) and hot gas bypass requirements, to name a few. This paper takes a practical approach in discussing the compressor and process parameters necessary for building these dynamic "high-fidelity" industrial-compressor models. We identify compressor inputs and compressor responses that are faithfully modeled by first-principle equations available in the simulation software and those that typically require a compromise between an "ab initio" and data-fitting approximation. We discuss the simulation's tendency to overstate pressure excursions during surge events and understate the compressor operation in the "stonewall" region. We also discuss using the simulator software's compressor-stage enthalpy calculations to predict and quantify the compressor train reverse rotation. We use our broad experience and understanding of the compressor operation and simulation and our experience with the AVEVA™ Dynamic-Simulation "OTS" simulation software as the basis for this discussion. |
|---|