Sustainable Automation Through Energy Monitoring and Process Control

Image Source: Andrey Popov/Stock.adobe.com
By Frederick P. Fendt, Industrial Energy Efficiency Consultant, for Mouser
Published October 30, 2026
Industrial automation has always been about production. For decades, throughput, uptime, quality, and labor efficiency drove every investment decision, and energy costs were treated as a different category entirely. Most plant managers thought of energy costs as fixed overhead and the price of running a plant, not something that could actually be adjusted. When the utility bill arrived, the organization paid it and moved on.
That assumption is now being challenged. Investors, customers, and regulators want measurable reductions in energy consumption and emissions. The goal is to achieve those reductions without sacrificing the production performance and profitability that the business depends on.
The operational response to this challenge is sustainable automation, which entails applying sensors, controls, and process data to reduce energy waste as a deliberate objective, rather than a byproduct of improving throughput. Sustainable automation tools include submetering and flow measurement, variable frequency drives, monitoring platforms, and tighter control strategies.
My four decades of doing this work in chemical plants, paper mills, and industrial facilities have taught me that it’s harder than it looks. This article examines how manufacturers can make automation more sustainable by questioning existing operating procedures, providing operators with accurate, real-time energy data, selecting sensors that align with decision-making, and embedding efficiency gains into everyday plant operations.
Automating Inefficient Processes May Lock In Energy Waste
Automated facilities with poor energy performance tend to share a common pattern. The automation was built to replicate what the plant was already doing, just without a person standing there to do it. The automation delivers consistent, reliable execution of the documented procedure. However, engineers may fail to determine whether that procedure is appropriate for automation.
Too often, engineers talk to operators, document the existing procedure, and build a control system that executes it reliably. While this might seem like a sound approach, it skips the harder question: Does this procedure use more energy than the process actually requires?
A large petrochemical facility in the UK had deployed advanced process controls across the site and contacted me about a troubling problem: Every time the automated system took control, it shut down the whole plant. After investigating, my team found that the system had determined that the incoming raw materials were worth more on the market than the finished products the plant would produce from them. It determined that shutting down production and selling the raw materials was the economically correct decision. The calculation was working as intended; what the team had not anticipated was where its logic would lead. The experience showed why engineers must examine the assumptions and possible outcomes built into an automated strategy before allowing it to control the process directly.
The same thing constantly happens with energy decisions. For example, if a combustion system is automated without anyone questioning the efficiency of existing air settings, it will run at those settings indefinitely. A pump sized for a production load that no longer exists will continue running at full capacity. The automation does exactly what it was designed to do, but no one asked whether it should.
Bringing Energy Intelligence into the Control Room
After decades in and around manufacturing plants, I have learned that nearly all operators genuinely want to do a good job. They will make every effort when they have a clear path. In almost every assessment I have performed, the limiting factor was access to the right information. Operators can see temperatures, pressures, and flow rates, but the energy cost of their choices is invisible.
Under production pressure, operators fall back on settings that feel safe and familiar, widening combustion air margins, raising pressure setpoints, and leaving equipment running through changeovers. These choices aren’t irresponsible—they are rational given the information available.
The solution is to make the energy cost visible alongside the process readings. At one facility, my team added a background calculation to the distributed control system (DCS). It evaluated operating conditions in real time and displayed a prompt on the operator’s screen whenever a configuration change could lower energy use. If conditions favored switching a large pump from an electric motor to a steam turbine drive, the screen would show the recommendation and the dollar value (e.g., switching would save $263 per hour under current conditions). The operator still made the final call; there could be maintenance, reliability, or production reasons to retain the electric motor. But now they were making that decision with the exact cost in front of them.
Identifying Hidden Utility Losses Across the Plant
Across dozens of our assessments, the same finding held true. At facilities that had not conducted a steam trap survey, compressed air audit, or insulation check in the past year, roughly 10 percent of the energy bill sat in losses that nobody was counting. Steam leaks and failed steam traps were present at nearly every site, along with degraded pipe insulation, compressed air losses, and nitrogen escaping through open tank vents.
A failing steam trap does not trigger an alarm, nor does a stuck tank vent. These losses accumulate quietly, year after year, until they are just part of how the plant operates. The pattern shows up in the field regularly. Walk through a facility on a cold morning where steam losses are thick enough to cut visibility across the yard, and the plant manager will tell you they run regular steam trap audits.
Part of the reason these losses persist is organizational. Nobody gets a big promotion for finding a nitrogen leak. Plant managers will say it matters, and they mean it. But it doesn’t move careers the way a new process line does. In plants that don’t actively track these losses, nobody really goes looking.
Compressed gases are easy to miss because they do not look like energy. For example, most industrial plants use nitrogen across several applications. It keeps oxygen out of storage tanks, purges process lines, and pressurizes equipment. Compressed air drives pneumatic valves, instruments, and tools. The energy cost does not show up where the gas is used. It was paid upstream, at the compressor or the separation unit, and by the time the gas reaches a tank or a valve, that cost is long since spent. A stuck-open vent on a nitrogen storage tank appears to be a maintenance issue. Most facilities do not have anyone regularly checking tank vents, which is why they can run open for months. Every cubic foot vented to the atmosphere discards everything the separation and compression process consumed to produce it.
To combat these hidden problems, companies can install a monitoring system. A stuck-open vent now registers as abnormal flow demand, or a degraded steam trap shows as a pressure drop in the wrong place, giving maintenance teams a specific place to look.
Accurate Measurement Is Essential for Energy Optimization
You cannot improve what you cannot measure, and measurement quality varies enormously from plant to plant. Operators and engineers need to know which line used the steam, which compressor supplied the air, and which furnace operated outside its expected range. A site-level consumption figure tells a plant manager how much energy the facility used overall, but nothing about specific sources of waste.
The deeper problem is that many plants are showing estimates that are being classified as measurements. A typical steam distribution system has one accurate meter on the main supply line, but the branch lines feeding individual buildings and production areas often go unmetered. To calculate the flow, an engineer probably once estimated the percent of use (e.g., building A uses roughly 40 percent, line B about 25 percent) and recorded them in a spreadsheet. That spreadsheet has been generating the dashboard numbers ever since. The precise figure a manager sees for a specific production area may have nothing to do with actual flow. Instead, it may be a fixed percentage of the main meter reading, based on an estimate that may be years or decades old.
As a result, operating decisions rest on data that do not reflect physical reality, and engineers who invest genuine effort in efficiency improvements see no change in the reports. They fixed the leak, retuned the system, and documented the project, but the spreadsheet still shows the same allocation. The process had improved, but the reporting system never captured it. Eventually, the disparity causes engineers to stop putting in that effort.
The Right Sensors Drive Better Energy Decisions
Sensor specifications matter more than many engineers expect. A sensor that works well for periodic emissions reporting may be wrong for closed-loop process control.
For example, boilers and process heaters need a precise fuel-to-air ratio. The combustion system requires sufficient oxygen for clean, safe burning. But excess air wastes fuel by heating gas that contributes nothing to useful output.
The trade-off is driven by temperature. Reducing excess air raises combustion temperatures. At higher temperatures, carbon monoxide (CO) oxidizes more completely, so CO emissions fall. But those same elevated temperatures accelerate the formation of nitrogen oxides (NOx). Running the boiler with excess air does the opposite: Lower combustion temperatures suppress NOx but also reduce CO oxidation, increasing CO risk. The result is a single temperature curve that pushes CO and NOx in opposite directions, and no position on that curve simultaneously minimizes both.
Additionally, emissions regulations create a specific bind. State regulators may test a plant’s boilers at several operating conditions, measure the results, and set permit limits based on the facility’s best historical performance. The logic seems reasonable. If you’ve achieved it before, you should be able to sustain it. The lowest CO reading on record becomes the CO ceiling, and the lowest NOx reading becomes the NOx ceiling.
The problem is that those two best results almost certainly came from different points on that temperature curve. When CO was lowest, NOx was often highest, and vice versa. There may never have been a single operating point at which the plant achieved both simultaneously. The regulator built the permit by combining each limit’s best historical result into a target that the plant may never actually have demonstrated. Achieving both optimum results requires real-time combustion control precise enough to find and hold a very narrow window.
Measuring stack oxygen seems like the natural solution, and emissions regulations typically require it. However, the standard extractive method draws a sample from the stack, conditions it to remove particulates and moisture, and then measures oxygen content. In my experience, three to four minutes typically pass from sample extraction to a reading at the control system. That lag is fine for a compliance record, but not for a combustion system that can develop an unsafe condition in seconds. Sensors that measure oxygen directly in the exhaust gas stream, rather than in an extracted sample, return readings in seconds and are the correct selection for closed-loop combustion control.
The same logic applies to every energy-critical system. Specify the sensor for the decision it needs to support, not for what is already installed or what reporting requires.
Turning Energy Improvements into Long-Term Results
Energy improvements don’t sustain themselves. Without measurement, visible metrics, and management attention, efficiency gains erode. Operators drift back toward familiar margins, calibration slips, and control loops that were tuned carefully at startup gradually lose their precision.
Boiler blowdown illustrates how this happens. This process removes water from the drum to control the concentration of dissolved solids. If too little water is removed, solids begin to concentrate, creating reliability and safety problems. Removing too much water leaves insufficient hot water at operating pressure, thereby carrying heat out of the system that must be replaced by burning more fuel. It’s a balance that requires active management. Without regular review, optimized blowdown rates drift back toward excess, and energy costs rise again without triggering any alert.
Sustainable automation requires building in those review mechanisms from the start. That means baselines, alarms, maintenance triggers, calibration schedules, and dashboards that provide regular updates on energy performance to the right people. The goal is to make energy efficiency part of how the plant normally operates, not a project that gets funded once and then fades.
Sustainable Automation Depends on Visibility and Control
The most significant benefits of sustainable automation rarely come from automating existing procedures more efficiently. They come from determining whether those procedures should change first. That means finding where the energy consequence of an operating decision is invisible and building the system to make it visible. Measurement shows where energy is being used. Control changes how the process runs. Operating practices, built into the automation, keep the improvement in place.
In my experience, the practical barriers to process-level measurement have decreased considerably over the past decade, and what sustainable automation can realistically deliver has expanded with them. But real constraints remain. Emissions frameworks create trade-offs that automated systems cannot resolve without experienced operators, and measurement at most facilities is not yet close enough to confirm whether efficiency gains have held.
For engineers designing or modernizing these systems, the question worth asking at every stage is direct: Does the operator or control system have a real-time view of the energy cost of each available configuration? Where the answer is no, that missing visibility is the starting point. Specify the measurement, verify the sensor is accurate and fast enough for the decision being made, and connect the data to the process condition that drives the energy use.