Let’s discuss the significance of simulation software in production research and experimental research
Thread Content
This post was last edited by zhangyia on 2012-5-27 09:52. The members of this forum include those who are involved in design, production, and experimentation. The importance of simulation software for design research goes without saying. Aspen’s training materials come with standard answers. However, the importance of production research and experimental research has always been unclear. For example: What can simulation calculations do for a well-functioning production facility? (1) Things that can be calculated can generally also be measured. Why calculate anymore? What is the meaning of calculation? (2) When it comes to guiding production, how does it compare to advanced control? Good advanced control systems can operate directly and continuously online, replacing the daily tasks performed by operators, and enable gradual optimization while in operation ； (3) Guiding technological transformation falls within the scope of design, not that of production research ； For another example, I need to conduct an experiment – is simulation software useful at this time? Of course, I won’t do things that simulation software can calculate directly. Feel free to discuss the above two aspects; this can be done from a broad perspective of systems engineering, or by referring to specific devices or examples. Those with well-supported arguments will receive 5~20 points from the moderators. Everyone (including moderators) can also refute arguments; those who provide well-supported refutations will receive 5 to 20 points as a bonus.
This post was last edited by songpine on 2012-5-28 09:25. The essence of simulation lies in using theory to guide practice. (1) It’s true that many physical quantities can be measured in chemistry, but there are still many others, such as enthalpy and entropy, which are difficult to measure. It’s not possible to measure every physical quantity, after all For example, in a pipeline, you can measure temperature and pressure. But the components, compared to others, are not easy to measure. Especially those at the dew point – what’s the point of taking a temperature reading? The advantage of simulation is that you can obtain the parameters you need without having to perform many measurements. (2) When you need to perform optimization, it is often the combined effect of multiple variables. Simulation helps you understand the principles behind the impact of various variables, which in turn enables you to target your efforts more effectively and narrow down the range of control parameters. Rather than relying solely on advanced control systems to carry out optimization in a search-like manner. (3) If you are an operator and don’t want to improve yourself, then you really don’t need simulation. But if you are the plant manager, through production experience and combined with simulation development, you can improve the process packages, or even resell them. Isn’t this meaningless? At least some petrochemical plants I’ve worked with have done this. Even in the worst case, if there’s a problem with your device, can you diagnose it solely based on experience? (4) Speaking of experiments, do you conduct comprehensive experiments? Can you tell if your interpolation method is accurate? The purpose of the experiment is mainly to calibrate the simulation parameters in order to improve the accuracy of the simulation. As far as I know, the development of a process abroad involves thermal simulation, sensitivity analysis, dynamic simulation, computational analysis and design of various equipment and pipelines, pilot testing, pilot plant testing, full-scale testing, and finally commissioning of the process. Then guide production. It can be said that all of them rely on various simulations.
Let me help summarize upstairs: (1) Calculating things that cannot be measured； (Can we continue discussing the significance of unmeasurable variables for production?) ) (2) Optimization of process parameters ； (Advanced control isn’t like looking for a needle in a haystack, dear. Frequent adjustments to operating parameters mean that it’s impossible to rely on operators; instead, controllers are necessary. (3) Improve the process package for fault diagnosis ； (I personally strongly agree that simulation calculations can be used for fault diagnosis, which is not mentioned in Aspen’s training standard answers.) (4) The summarization is not good. Finally, the development of a process abroad always follows the steps of pilot testing, pilot plant testing, full-scale testing, and then commercial production; even today, most processes still follow this approach. When simulation tools became available later on, it simplified the development work to some extent. Of course, this is essentially a design issue.
It’s written really well; the gap between practical application and simulation is inevitable
I think this issue can be reduced to whether calculation is needed for production and experimentation. Because the essence of simulation is mechanism and model calculation. This question is based on accurate simulations; otherwise, there is no need for discussion. As for the gap between simulation and reality, it is necessary to identify and address that gap as much as possible, rather than denying it. Only in this way can the calculation results be used for optimizing processes and performing fault diagnosis. Another control process also requires attention to response time, as practical optimization often comes at a certain cost.
There are things that one can only truly understand their intricacies by using them personally and reflecting on them in depth.