The Possible Downsides to Excessive Automation

It has its place in the water systems market but should be applied only when needed.

By Ed Butts, PE, CPI

The July and October editions of my other column, The Water Works, presented an introduction to programmable logic controllers (PLCs), and the magnificent contribution these and other devices make to modern control systems.

Although sophisticated SCADA systems and PLCs generally provide a more reliable and accurate method of control through automation and a viable alternative to the old-fashioned method of hard-wired circuits using relays, timers, and switches, they are not without their potential drawbacks.

In this column, as a quasi-counter argument to earlier editions of Engineering Your Business, I’ll outline some of the limitations and dangers of modern control technology and warnings on when, how, and why they should and should not be used. In other words: Not all advanced technology is needed or warranted for every water system. Let’s find out why.

Some Personal Background

Although I generally practice civil, environmental, and water resource engineering, I also work with and am registered in the branches of electrical and control systems engineering.

Most of my control systems engineering work is dedicated to single unit water and wastewater pumping systems, but I also design control systems for industrial processes; multiple-unit water and wastewater pumping systems; geothermal, dewatering, and treatment systems and plants. This is all in addition to supervisory control and data acquisition (SCADA) systems, often referred to as telemetry, that I design for municipalities, industrial, and irrigation applications.

I began this work during the late 1970s and have witnessed a technological explosion that began with simplistic analog circuits and hard-wired discrete (on/off) control systems to our present use of digital technology.

During my first 10 years of practice, we were content to utilize analog circuits, usually 4 to 20 milliamps, for transmitting variable parameters including reservoir water levels, flowrates, and pressures to a remote site using Frequency-Shift-Key (FSK) transmission methods.

FSK is a frequency modulation method in which analog or digital information is transmitted through discrete frequency changes over a carrier signal. The carrier signal is generally transmitted through a leased telephone pair or dedicated conductors. In FSK, two carrier signals are used to produce FSK modulated waveforms. These are called “mark frequency” and “space frequency,” often shortened to generically as mark and space.

A modulator transmitted the signal and a demodulator on the other end received and translated the signal. This system was reasonably reliable, easy to understand, install and troubleshoot, and relatively inexpensive—but it was limited as to the reliable transmission distance.

Today, we use much more sophisticated communication methods in addition to telephone or hard-wired circuits—including FM radio, microwave, cellphone, and even satellite in some cases.

Another addition to water system control has been the advent of programmable logic controllers (PLCs). The earliest versions of commercially available PLCs were introduced during the 1970s.

My first exposure to a PLC was with the Texas Instruments (now Siemens) model 5TI unit. This was truly the most basic PLC at the time. It was a tower unit with all 20 discrete inputs and outputs (I/Os), consisting of 12 inputs and 8 outputs, built directly onto the processor. It was capable of multiple operations including math functions, counting, analog (with a separate module), comparative, and timing capabilities.

In this case, I applied the PLC to operate a leachate land disposal system that required blending of fresh water from a well to lower the salinity of the applied solution by dilution to prevent an overapplication of salts onto the irrigated land. The system used an online conductivity sensor to monitor the water’s concentration of salts with a dual solenoid pulsing control valve to vary the amount of dilution water introduced into the flow stream to maintain a tight band of conductivity.

Following an initial adjustment and proving period, the system worked well and performed the required function with minor readjustments and downtime for several more years.

An Example of Possible Downsides to Excessive Automation—Boeing 737 Max

Now to some of the potential downsides of wrong or misapplied automation. For the purposes of illustration, I will cite the recent example of Boeing’s 737 Max airliner. The 737 Max jet airplane is the fourth generation of Boeing’s popular 737 airplane and was introduced to great fanfare in 2011 as an update and replacement to all past versions of the iconic and most common Boeing plane.

It began its maiden flights and went into service in 2016. Subsequently, the passenger airliner was grounded worldwide between March 2019 to December 2020 after 346 people died in two separate crashes that occurred within five months of each other.

Both crashes were attributed to recurring failures in the new Maneuvering Characteristics Augmentation System (MCAS) consisting of a new control system and the sensors that monitor the plane’s “angle of attack.”

The angle of attack refers to the relative angle between the airflow under the wings and the airplane’s orientation. An excessive angle of attack, as compared to the plane’s forward speed, will disrupt or impede the air traveling under the wings, leading to a loss of lift.

This culminates in a condition known as a stall. In these instances, the sensor transmitted the wrong information to the MCAS, indicating the angle of attack was too severe and approaching stall speed where the plane’s forward speed cannot support the weight of the aircraft and sustain continued flight.

The system then pitched the nose down to correct the angle of attack and increase air speed, but the pilots realized this was a wrong command and they tried to override the system, but they were unable to override the command. Both planes were sent into sharp dives the pilots could not recover from and the planes crashed.

The same type of situation occurred in October 2008 when an Airbus model A330 jetliner encountered a similar software glitch. The flight computers had misinterpreted the angle of attack and altitude input data, causing the aircraft to pitch suddenly downward, instantly dropping 650 feet from its intended flight path from Singapore to Australia. Only the fast actions on the part of the pilots allowed recovery of the plane and averted a potential crash.

These are extreme examples with hundreds of unfortunate and potential fatalities where an advanced control system, intended to make the airplane and air travel safer, was introduced and used before all possible hidden glitches in the system were anticipated, discovered, vetted, and corrected.

In these cases, the primary issue I have with the use of this excessive technology is that it allows pilots to more heavily rely on the advanced computer technology to operate the flight controls while forgetting or ignoring the basic human aspects and responsibility of flying the airplane. Therefore, when human intervention is needed, the pilots may lack the fundamental skills, experience, and training necessary to perform this required function.

Another Example of Excessive Automation—This Time from a Water System

Another real-world example of when excessive automation may present operational issues involves one of my municipal clients. This client provides water and wastewater services to a city of more than 100,000 people, using several zones and booster and wastewater pump stations.

For decades, the city was able to operate their pumping stations for both applications, using fundamental types of control systems and basic remote monitoring capabilities with a predictable number of failures and anomalies.

However, in response to the wider industry use of PLCs and more sophisticated control systems used in pumping systems during the early 1980s, the city began to request more advanced methods of control for new pump stations, as well as retrofitting many troublesome existing facilities.

There was a plethora of system integrators, PLC salesmen, and firms lined up to assist with this effort. This resulted in the widespread implementing of PLCs with more sensors, data collection points, and greater operational logic, resulting in more I/O spaces and demands.

As with many new technologies, the initial use of digitally based control logic was met with resistance from many of the old-time operators and engineers. But the decision makers were convinced that modernization was necessary to avoid being left behind in the digital world.

The PLCs largely replaced the former conventional technology with little, if any, control redundancy added. The new control system mainly operated using remote supervisory control with minor local input by transmitting the data from the remote site to a master site and then receiving the command from the same master site.

Soon, many of the systems were inundated with overflow alarms; pump failures; low or high reservoir levels; incorrect water level, pressure, and flow readings; erroneous or false commands or valve positions; and other operating anomalies due to improper or inadequate design and vetting of the control system, communication failures, and failure to use adequate local backup controls.

This was a case of overreliance on a new remote-control method without considering the potential problems that can occur in actual use of the technology due to communication failures or incorrect programming or data. Enhanced control redundancy, generally through the addition of parallel or backup conventional local controls, was soon implemented resulting in more reliable control and less system failures.

The Connection to Water System Control and Use of Variable Frequency Drives

You may justifiably question the parallel with the controls used to operate a jet plane versus a water system at this point, but they are undoubtedly there. And please don’t misconstrue my message—no one welcomes and implements the enhanced accuracy and reliability of automation more than I.

It’s just that control and water system designers must recognize the potential limitations and pitfalls of excessive automation and incorporate adequate safeguards along with some redundancy and flexibility into the system design.

Although the use of PLCs, variable frequency drives (VFDs), and similar controllers generally provide a cost-effective and easily expandable alternative to hard-wired controls, incorrect equipment selection or programming and excessive reliance on automation systems can present serious challenges and potential failure modes to control, and in turn, water system reliability and functions.

In my judgement, one of the most common violations due to an overdependence on automation is the too often use of a VFD when it is neither indicated or warranted. From an engineer’s perspective, there are four main reasons to invest the funds needed to use a VFD over standard motor control methods:

  • Energy efficiency, which leads to improved economy and lower operating costs
  • Lowering pumping flowrates and heads to match actual system demands in a water system control mode
  • To reduce the starting inrush and operating stresses on electrical systems and pump and motor rotating components
  • For other strictly system operational purposes when needed.

These justifications are not necessarily interdependent—that is, using a VFD to lower the stress on a pump’s rotating components during starting or while operating may not automatically result in an equivalent savings of energy.

Because VFDs represent a significant investment, it is incumbent upon the system designer to conduct no less than an offhand evaluation to determine if the addition of a VFD is actually worth the capital and life-cycle costs—not because it’s the new thing to do.

In addition, many VFDs are equipped with the capability, if not the actual hardware, of a PLC as an element of the programming interface. This provides greater flexibility and enables the programmer to insert minimum and maximum motor speeds, usually as a function of frequency, in hertz, overload settings, speed rejection parameters, and in some cases, an underload adjustment.

A speed rejection adjustment is a particularly valuable addition to a VFD to prevent sustained operation at a pump’s critical speed, important for most deep well pumps. The application of a VFD when it is not needed or entirely functional, such as with a flat curve pump, will generally represent an unnecessary expenditure that will end up operating as a glorified reduced voltage starter.

This is the point when water system designers must carefully examine the system operating hydraulic parameters, including variable flowrates and heads versus probable pump and motor speeds, and select a VFD only when warranted by the actual system requirements—not simply as a revenue source. As good as they are, VFDs definitely have their valid place and use in the water systems market and should be applied as such.

Until next month, work safe and smart.

Learn How to Engineer Success for Your Business
 Engineering Your Business: A series of articles serving as a guide to the groundwater business is a compilation of works from long-time Water Well Journal columnist Ed Butts, PE, CPI. Click here for more information.

Ed Butts, PE, CPI, is the chief engineer at 4B Engineering & Consulting, Salem, Oregon. He has more than 40 years of experience in the water well business, specializing in engineering and business management. He can be reached at epbpe@juno.com.