What are the key components of a successful process control and optimization strategy?

First: What is the business case? pain point? opportunity? Who are the stake holders? Do the stakeholders believe in the opportunity? Understand that. Second: Gather direct data (process measurements), indirect data (operator actions), stakeholders feedback and validate everything around the business case Third: Design a solution that suits the problem and validate the solution with the end users and stakeholders. Do they believe in the solution? or they are pushing back? Use judgement before proceeding Then you implement the solution and tweak it to work better.

La recopilación de los datos es un paso critico, debe garantizar que la información obtenida es una muestra real del proceso que queremos analizar. De no ser así, podríamos tener lo siguiente: - datos falsos - comportamiento no real del proceso - invertir tiempo extra en la limpieza de los datos

Safety of people, employees should work in a safe environment this means that the process control system is called upon to achieve the safety of employees and process unit equipments in general the plant, quality of product, the quality of the product should confirm to customers satisfaction, at the same time the quality of the product represents the value of money and maximum Production can only be achieved by good process control system and continuous process improvements

During the data collection and analysis stage it is crucial to understand what the data represents using process and operational knowledge. What parts of the plant were in closed loop? What was in open loop? Was the process at a normal operating condition or at an abnormal one? If the data that is needed to create useful models or understand the full range of the process, then it is necessary to generate it either by offline modelling or by performing plant step tests.

This article seems to confuse and to put in the same bag two different but complementary things: the implementation of a baseline control system such as DCS/SCADA and the implementation of the APC or optimization layer. Especially for the baseline DCS/SCADA, the objective of safety and (cyber-)security of high importance.

Before implementing more advanced control strategies like APC or RTO, it is important to do a thorough review of the base regulatory system. The base system control loops performance should be checked. Any re-tuning or necessary maintenance on any control valves should be done prior to implementation of an APC. It is also important to understand how the APC controller will communicate with the DCS system.

It is clear that control systems are really helpful to monitor and control system better, but from my point of view it is necessary that operators have enough information about the system and do not rely on it to work. I suggest that they control process manually from time to time. It helps them to increase their knowledge about that particular action.

A lot of great comments and the article does touch on the key factors. It’s a little unclear as it seems to talk about the system capability before the design. Basically saying you need to have a system that can do everything! But it doesn’t really address the essential question - what is the system requirements? Not every control system needs fuzzy logic and optimization. All those capabilities typically mean higher price. So I would just add - just because you can do something doesn’t mean you should. I would say focus on the requirements. Select a control system that is built to handle more complex control in case you need to add it in future. But make sure you don’t spec what’s not needed. Or your project may get squashed.

A good control system is indeed the base to a stable and safe system but there good alarm management may not be forgotten. Loads of alarms during normal operations may cause the operator to ignore them and miss important alarms during upset situations. Good instrumentation is indeed also needed cause the controls may be good but if they're in manual because of failed instruments the controls can't help. And don't forget that your control system may also need updates to keep working.

This is by far one of the most important steps in a successful implementation of an APC or RTO. If operators do not understand how the tool works, they will not use it or not fully utilize it. It is crucial that control engineers do a good job explaining how the APC or RTO works with the underlying control systems such that operators understand the goal. Operators and process engineers should always be involved in the early design discussions for an APC or RTO application because they have insight that most control engineers do not possess. It is important that the operator training is regularly repeated.

Operator training needs to consider how information presented is perceived. I have worked in a facility that included 3 languages and several levels of illiteracy across the operator spectrum. Which inputs and displays will produce a correct response

What an important item in this article. I strongly believe that having skilful operators ARE GAME CHANGER. Therefore, it is recommended that we consider to cross training.

Given that the development of an APC or RTO application is a time consuming and rather costly process, it is important to not waste this effort by not doing proper follow up of the tool. Depending on how often the process changes, regular review periods should be established. The benefits of the tool should be established typically one year after use to compare the plant's operation before and after the install. For a common APC implementation like Model Predictive Control, the constraint limits should be reviewed regularly during daily optimization/operational meetings to ensure all benefits are captured.

Performance monitoring should also be used to look for maintenance issues/ opportunities. Valve performance degrades over time, same for pumps, motors, etc. If you integrate condition monitoring into the system you can increase uptime and reduce unplanned maintenance.

its important to connect the end with means to ensure the initial desired outcomes and the reason for investment is achieved.

Often advanced control loops remain off because the lower traditional PID do not attend performance expectations or have hardware problems: instrumentation or control valve. A good PID tuning and up-to-date maintanance of instruments and control valves turn possible to return PID loop in auto and then run the advanced control.

Deep understanding of processes are invaluable assets to proper controls design. A shared understanding and mutual respect between the operations and the controls team is fundamental. Process which are highly nonlinear become difficult to control and thus system identification is the first step of any controls study. Once done, we start developing basic controls like FF, FB based on these studies. Currently, PID control is the first form of control to be explored. PLC design, protocol matching and deployment configurations are important in this aspect. Once completed, a CBA of APC is done and we further try to implement advanced strategies like MBC, estimators which are fall under the optimization domain.

There are 2 parts to this at a high level: - Optimise what you have:this looks at the combination of input variables and what they combine to give as an output that is consistent (max output, minimum variation). This is fine and will typically give 5% to 30% improvement annually - Understand and break constraints: this looks beyond nominal performance of the input variables and considers what can be done with and without CAPEX to address constraints. This will typically give 30% to 100% improvement Whereas the first approach requires a statistical approach to a large data set (some argue can be done by AI/machine learning), the latter requires a deep understanding of science and engineering, out of box thinking, strong goal orientation