How a Three-Shift Plastics Plant Tied Every kWh to Output and Opened a New Revenue Line
For a mid-size plastics manufacturer running three shifts, energy is one of the largest lines in the cost of goods sold. The presses run around the clock, the chillers never rest, and the compressed-air system feeds nearly every station on the floor. Yet for years the plant treated all of that as a single utility bill, a number that arrived once a month and could not be pushed back on. Nobody could say what a given part actually cost in kilowatt-hours, or which machine was quietly bleeding money. This is an illustrative account of how a plant like that changed the picture with EnergyOS, the managed submetering and sensing platform from Emergent Metering.
The core shift was simple to state and hard to achieve before: energy stopped being a monthly mystery and became a per-unit number the plant could manage shift by shift.
The operator
The plant makes injection-molded components for automotive and consumer goods customers. Three shifts keep roughly two dozen presses busy, supported by a central chilled-water loop, a bank of air compressors, and the usual mix of conveyors, dryers, and material handling. Margins are thin and contractual, so a few points of waste on the floor matter to the year-end number. The operations team was capable and experienced, but they were flying with one instrument: the utility meter at the property line.
The challenge
Everyone knew energy was expensive. What they did not have was any way to connect it to reality. The presses, the chillers, and the compressed-air system all drew from the same feed, so a spike in the bill could not be traced to a machine, a shift, or a product run. Compressed air, often the single most wasteful system in a plant like this, was completely invisible. Chiller performance drifted without anyone noticing until a hot week made it obvious.
On top of that, unplanned downtime kept eating into output. When a motor or a compressor failed mid-run, the line stopped, scrap piled up, and a maintenance scramble followed. Management had also heard that local demand response programs paid facilities to curtail load during grid events, but with no visibility and no controls, the idea of committing to a curtailment felt like a risk they could not size. So they left that money on the table.
The deployment
Emergent Metering deployed EnergyOS as a managed service, which meant the plant did not have to build or babysit any of it. The first step was submetering. Instead of one meter at the property line, the platform built a metering tree that broke the load down by production line and by major equipment, with the compressed-air system and the chiller loop metered as distinct nodes. Virtual meters let the team roll individual presses up into logical groups, such as a customer program or a shift, without adding hardware.
With production counts fed in from the floor, EnergyOS built energy-per-unit dashboards. For the first time, the plant could see the kilowatt-hours behind each part, normalized against weather where it mattered and benchmarked line against line. A press that looked fine on a throughput report suddenly stood out because it burned far more energy per unit than an identical machine two bays over. That kind of hidden waste, invisible on a single bill, was exactly what the metering tree surfaced.
Machine health and faults
Metering told the plant where the energy went. Sensing told them how the equipment was doing. EnergyOS set machine-health baselines on the compressors and chillers using vibration and accelerometer readings alongside current-draw monitoring. Once a normal signature was established for each unit, the platform watched for deviation and ran fault-detection rules against the key equipment.
The payoff showed up when a motor on one of the presses began to run rough. Vibration climbed above its baseline and current draw crept up in a pattern the fault rules recognized. A real-time alert went out with escalation, so the right person saw it rather than a message sitting unread in a queue. Maintenance pulled and replaced the motor on a planned window instead of during a production run. Catching that one failure early avoided an unplanned stop on a three-shift line, where every idle hour is scrap and missed shipments. Directionally, plants that move from reactive to condition-based response on critical equipment tend to see a meaningful drop in unplanned downtime, and this deployment fit that pattern.
Demand response as revenue
With load finally visible and controllable, the demand-response question stopped being scary. EnergyOS enrolled the site in a local demand response program and used peak-demand tools, demand limits, event tracking, and forecasting, to model what the plant could safely shed and when. When a grid event was called, the platform tracked event performance against the commitment and handled the revenue accounting, so finance could see what each curtailment actually earned.
Curtailing load during grid events turned into a genuine revenue line rather than a hopeful guess. Because the platform showed exactly which loads could be trimmed without hurting active production runs, the operations team could say yes to events with confidence. Demand response revenue for a facility this size is typically modest against total energy spend but pure upside, money the plant was previously walking past.
M&V and the finance story
The plant had suspected its compressed-air system was leaking value, and the submetering confirmed it. After a compressed-air fix, tightening leaks and adjusting controls, EnergyOS ran a measurement and verification project to document the result. Using a weather-normalized baseline from before the change, the M&V project isolated the savings from the fix itself rather than from seasonal swings or production changes.
That mattered because finance does not act on hunches. A verified, defensible savings figure gave the CFO something to book and something to trust the next time operations asked to fund an efficiency project. M&V turned a floor-level improvement into a documented financial result.
The results
Taken together, the deployment moved the plant from blind to instrumented. Hidden waste was found on specific machines rather than assumed across the whole floor. Energy-per-unit improved as high-cost presses were tuned or rescheduled. Downtime fell because a failing motor was caught before it stopped a shift. Demand response opened a new revenue line the plant had been ignoring, and the compressed-air fix came with verified savings that finance could stand behind. None of these are audited figures, but each reflects the typical, directional gains a plant of this profile sees when energy becomes a managed number instead of a monthly surprise.
Sensing hardware is powered by Monnit, delivered inside EnergyOS by Emergent Metering.
Ready to tie your energy to your output? Schedule a Platform Demo. Call 215-645-7141.