The definition of Advanced Process Control (APC) is ever changing with the advances in the microprocessor and control technology. The advent of the distributed control system (DCS) in the 1970’s, and its wide use in the refining and petrochemical industry, paved the way for easy implementation of the existing control techniques such as cascade, feed forward, non-linear level control, Smith predictors, constraint control and even decoupling control. These were control schemes that went beyond the PID (Proportional-Integral-Derivative) single loop feedback controller and provided distributed and supervisory control. It is practical to say that while regulatory controls maintain or close the mass and heat balances, advanced controls manipulate the mass and heat balances to achieve the best performance or quality.

By the 1980’s, a new type of controller was born that not only provided feedback and feedforward control, it took constraint control to new heights. The multivariable predictive controller (MVPC) would be most effective when complex interactions existed in the process. The controller is model based and predictive, and would provide feedforward, feedback and constraint management in a single controller. The computing requirements however, was beyond the capability of the available DCSs, hence the need for the employment of powerful supervisory computer systems connected to the plant’s communication link.

Having a powerful computer connected to the plant collecting real time data opened up numerous possibilities for the emergence of complementary technologies. For example, the inferential estimation technology that would infer the required composition of the stream, which is used in the feedback control, in the absence of on-line analysers or long delays in the measurement signal. The technology behind the models for these estimators range from simple correlations to first principles chemical engineering and neural networks modeling.

Another example of advanced control is the Expert Systems technology which captured people’s imagination back in the early 80’s. It is based on capturing knowledge of the best operators and transforming real time data into useful information through reasoning and analysis. The technology has had relative success in the oil refining industry with applications such as start up/shut down, emergency and abnormal condition monitoring.

With the leaps in the microprocessor technology in the 1990’s, and the need to have a unified and consistent interface for the engineers and operators, various DCS vendors embedded the computer applications and in particular, MVPC in the DCS hardware. In this decade, the MVPC application numbers grew to thousands and the question with these controller was not so much “does MVPC work?”, but ” how much benefit will it have for my application?”. The same cannot be said about the non-linear rigorous Real Time Optimisation (RTO) technology, which despite the fact that it has been around for more than 20 years, it has not been able to make the same impact. The early versions of this technology was based on internally converged models, better known as sequential modular optimisation. The units in the process industry require complex set of equations to represent an accurate model of the plant, hence, problems often lead to thousands of equations for reactors and distillation columns in a flowsheet. With faster computers, this problem was partially resolved, however, a new technology based on open equation modeling was also introduced which made better use of the computer memory by solving the equations simultaneously. The applications of RTO were primarily unit or multiunit optimisation and in very few cases applied to complexes and even fewer applications of refinery wide RTO. RTO is not a replacement for the Linear Programming (LP) which has been around for 50 years in the refining industry in one form or the other. Towards the end of the last century, every refinery utilised LP technology for business optimisation.

The purpose of this section is to provide the users with a background knowledge of the technology so they can make intelligent decisions about their existing APC and to embark on installing new ones with open eyes.

The diagram shows the plant operations pyramid. The pyramid as it stands represents the data transfer rates for the applications. The inverse of the pyramid is also true for the computations per solution.