Our profession has changed a lot during the last five decades. In the first edition of The Instrument Engineer's Handbook (IEH), I described how to tune single-loop pneumatic controllers and was perfectly satisfied with floats for level and filled bulbs for temperature measurements. At that time, in some plants the main job of an instrument engineer was to clean plugged sensors and fix stuck control valves. Our control panels were full of pushbuttons and manual-loading stations, showing that the designers trusted the operators more than automation.
Today, when we use multivariable envelope algorithms to optimize complete plants, when the control theory of our industries has spread from military to medical applications, and when we block operators from overriding or deactivating automated safety systems, those early years seem like distant memories.
Fifty years ago, process automation as a separate discipline didn't exist. I taught process control in the chemical engineering department of Yale University, and the first edition of the IEH came out of the electrical engineering division of its publisher. Why? Not because Yale or Chilton disliked our profession! They didn't know it existed! This has changed a lot.
Today in some industries, the roles have actually reversed. New developments in chemical or electrical engineering are relatively few, while the science of automation and process control is exploding. This increase is accelerated by the dropping cost of computer memory, by the standardization of protocols, by the ease of configuring complex algorithms and dynamic displays. This sudden rise in the recognition of our profession is due not only to the better tools at our disposal, but also to what we are doing with these tools.
It is now recognized that automation can simultaneously increase profitability and safety! No other profession can increase the GDP without building a single new plant. We can do that just by optimizing the existing plants. We can do that while also reducing pollution and energy consumption, solely through the application of state-of-the art automation. We can increase productivity without using a single pound of additional raw material and without spending a single additional BTU.
We can do that because, by replacing manual control with automatic, we maximize not only production but also safety. It is because delegating accident prevention to automatic systems increases safety, while delegating it to panicked, disgruntled or badly trained operators does not. For some, this change in the safety culture is hard to accept, and this group even includes some of our own older colleagues, who spent a lifetime designing manual or semi-automatic controls.
We now know, for example, that automatic flooding of the Fukushima reactors during the 45 minutes between the earthquake and the arrival of the tsunami could have prevented the meltdowns, and the automatic venting of hydrogen could have prevented the explosions. Similarly, lives would have been saved at the BP accident, if the platform was automatically disengaged from the well, or at the Asiatic jet accident, if landing was automatically aborted when the speed was manually allowed to drop below its safe limit. My purpose here is not to list cases, but to emphasize the need for changing the "safety culture" of our profession from manual to automatic.
Read Also: Automation Could Have Saved Fukushima
Automatic safety controls are the airbags of industry. The operator of the process doesn't need to actuate the safety controls, nor can he disable them, because the controls trigger the safety protection automatically when the situation requires it. In the coming age of cyber terrorism, the acceptance of this safety culture will be even more important.
IT and Automation
Another major step in the right direction was the separation of information technology (IT) and process control. It took some time for industry to realize that IT has no safety culture; that nobody ever got killed by an office machine, unless it fell on them. Another cultural difference is that process control is continuous, while IT runs during daylight. Therefore, today we know that process data can flow from the process to the enterprise (IT), but direct queries from the enterprise to the process control system are absolutely unacceptable.
In the distant past, it took decades to standardize on the 3-to-15 PSIG (0.2 - 1.0 bar) pneumatic signal and years to standardize on the 4 to 20 mA DC electronic signal range. Everybody benefited from these standardizations because they allowed the products of different manufacturers to "talk" to each other.
Today, we work with microprocessor-operated intelligent field devices (IFDs), digital transmitters and digital valve actuators. They communicate digitally, and can handle multifunctions, such as detecting multiple variables, and simultaneously carrying out multiple tasks. Similarly, the valve actuators can have a variety of algorithms that can act as valve-opening sensors, limit switches or as remotely tuned PID algorithms.