Is it time for you to optimize your operations?

I am planning a synthetic ammonia plant. I do feel that we are not doing enough in terms of production optimization. Something like this happened in my factory a few days ago: The above command issued an instruction to add a load of 1000kg/h, but the output increased very little in one day. It often even happened that the total natural gas consumption increased after adding the load, but the output did not change at all. This reflects a phenomenon that everyone's understanding of production optimization is not yet in place. Operators are just ok with ensuring that the device can run continuously for a long period of time. Production optimization has a long way to go.

Have a few questions? Have you done any evaluation of existing equipment and operating status? Have you ever done any operation optimization analysis? Have you issued operational indicators that have been adjusted after operational optimization?

Generally, equipment is in the exploratory stage for a period of time after the start of construction. At this stage, we accumulate a large amount of data and analyze the relationship between these data and the relationship between these data and product quality. Find the nearest operating point. Guarantee the quality of the product. For example, for the operation of propane and propylene in a liquefied gas separation device, it was found during production that with the frequent changes in the C5 content (0-5%) of the raw material, the temperature of the tower kettle also changed greatly under stable operating conditions. If it cannot be accurately determined, The reasons for continuous changes can easily lead to incorrect adjustments, resulting in tower fluctuations, which may affect the product quality at the top of the tower and the extraction rate of propylene. We performed statistical analysis on the production data and summarized the relationship between carbon 5 content and tower kettle temperature for operational reference. Under a certain carbon 5 content, we refer to this relationship diagram to control the tower reactor temperature at a certain value T. If the control temperature of the tower kettle is less than T, the extraction rate of propylene will be reduced. If the temperature of the tower kettle is greater than T, the top product will be unqualified.

Very good way to adjust a single variable! But what if there are more variables that need to be considered? (Like backflow. )

We are a small to medium-sized chemical factory. The hierarchical salary distribution method based on measurement is adopted. The boss sets a unit cost for each raw and auxiliary material and product, and wages = output wages + savings. And we strongly advocate small reforms. The unit consumption is often adjusted once every six months! In the past ten years, some indicators are close to theoretical consumption! Brainstorming, of course, the optimization of operations and the optimization of processes and equipment are indispensable! I won’t list them all here. What I want to say is the prerequisite for operational optimization: There must be a good incentive mechanism! ----Otherwise, how can we work hard? :lol :) :P

I would also like to share some immature opinions. The chemical production process changes due to changes in raw material composition, which affects production efficiency. And maintain the optimal value, adjust the raw material changes as little as possible, and find the best operating status for each position's safety under the premise of keeping the load as small as possible, and the final output will be higher and the consumption will be lower.； During long-term operation, stop the vehicle less often, deal with problems promptly, and do not delay the operation until it cannot stop for the sake of production. When the design capacity is reached and the load is increased, the increase in output will not be obvious, but the increase in consumption will affect the efficiency.

I think this requires accurate records during the normal operation of chemical production, determining the main factors affecting the return flow based on my knowledge and experience, and then making graph analyzes of the return flow and these factors. For example, on the premise of ensuring that the product quality is qualified, we know that the factors that affect the reflux flow include the reflux ratio and the feed amount. Then we can make a graph analysis of these two parameters on the return flow for operation reference. In scientific research, when there are many influencing factors, we can use orthogonal testing methods to shorten the workload. However, for production testing, we can only slowly accumulate production data and conduct careful research on it to find out the relationship between these factors and product quality.

If there are multiple variables, you can use the orthogonal method to do it.: First, fix a few quantities unchanged and change one of them. ； Then change another quantity and fix the others. . . . . . And so on. It’s just more time-consuming

Is it possible to use computer simulation to optimize operations? In this way, is it possible to find the optimal solution for multiple variables at the same time?