The MOD Squad: Process automation at Dow

In the early 1960s, Levi was R&D director and later general manager at Dow’s Texas Division. Under his direction, new processes and vast improvements to existing processes came into being. His words ring as true today as they did then: “If it doesn’t work, there is a reason why. If it does work, there is a way to make it better.” He embodied Dow’s passion for safety and operating discipline, and planted the technological seeds that brought them together. He didn’t know all that had to be done to achieve this goal, but he knew how to get people excited about the idea. The term he used to express his concept was “Operating Discipline.”

Levi shared his vision by traveling around the company to Texas, Midland, Sarnia, Canada and many other Dow sites worldwide. He didn’t just talk to control engineers, but to people at all levels, including process and instrument engineers, manufacturing, plant or line people, business leaders and senior management. Levi wasn’t the actual technology inventor, but his influence is easily seen in the early development efforts that began during this decade.

Converging Activities
Over the years, there were technology development activities throughout Dow, each addressing a different aspect of the overall automation opportunity. These activities ranged from creating process information systems to the developing closed-loop control applications. Ultimately, they all converged to one set of systems, but this took several years. Information sharing between engineering teams at multiple sites wasn’t what it is today. Travel between sites was limited, and the technology that facilitates information exchange between sites and work teams that we now take for granted today wasn’t a reality back then.

MOD 1 is Born
One major activity began about 1968 in the Midland Analog Simulation Facility, which was seeking better production process control. By performing differential equations on analog computers, researchers found they could simulate any conceivable control strategy. As a result, analog simulation was widely used to perform studies that would determine the best control strategy for controlling a unit operation or an entire plant.

However, the development of analog simulation also led to some frustrating times. Existing hardware, such as panel-mounted conventional controllers, often became a barrier to implementing the best process control methods. Compromises also had to be made to fit existing hardware.

Despite these constraints, some process control ideas were implemented on industrial analog computers built to withstand hostile environments in the field. However, while they had the flexibility to implement control ideas without hardware barriers, analog computers were costly and very limited in capacity. The logical question was: how to implement the technology of a large analog computer in the field at a much lower cost to the company?

The answer was MOD 1. Started in 1969 on a batch distillation column, MOD 1 resembled an analog computer, and had similar logic and computing components. It had approximately one-quarter of the power of the present MOD 5 computer, but it was the foundation for sequencing, alarms, flexible analog and digital control—the basic ideas for automating a chemical process. Built by Taylor Instrument Co., which is now part of ABB, the first MOD 1 cost $85,000.

FIGURE 2: THE MOD SQUAD GETS DISPLAYS
The MOD Operator Station provides a standard graphical interface to the process control system as well as an integrated environment for development.

This Operating Discipline philosophy was implemented using the “sequencer” concept. The importance and value of subdividing the entire plant into unit operations (sequences), and then subdividing the time up into steps for each sequence boggles the mind. Plant people could now think about, document, and automate all of the actions required to operate a plant, including start-up and shut-down. The control system would then provide relentless enforcement of the best way to run the processes.

These ideas often met with opposition because they were viewed as impractical or unachievable. The trench-by-trench effort to get the plants to accept and use this level of automation was fought for many years, and is still being fought in some instances. However, the favorable economic impact of the successful implementations were felt throughout our company.

MOD 2 Moves to Mode Modular
By 1970, MOD 2 was created to reduce costs, and move from a standalone to a mode-modular system. It was the first process control system with a standard system packaging method. MOD 2 also used digital modules from Digital Equipment Corp. and hardwired logic cards. However, MOD 2’s analog function modules were designed internally.

MOD 3 Increases Control Capability
Coming into general use in 1972, MOD 3 was developed to increase system control capability and reduce overall cost. Some rewiring was required to change the control logic. Installation of a MOD 3 in a Latex plant in Sarnia, Ontario, was the first use of MOD’s concepts outside the Michigan division.
MOD 3 used the first all-digital CPU, which was called the Alpha 16 or the Naked-Mini. This CPU was made by Computer Automation Inc., and had 4 KB to 32 KB of memory. MOD 3 also was the first process control system to use DOWTRAN, a computer language for end-user programming.