Industrial Submetering: Turning One Energy Bill Into Actionable Information

The total at the bottom of an electricity bill can tell a manufacturer how much energy the site has bought. It cannot explain which production line used it, whether demand rose because output increased, or why consumption remained high long after the last shift ended. By the time those questions reach the monthly review meeting, the detail needed to answer them has usually disappeared inside one site-wide figure.

Industrial energy submetering fills that gap by measuring selected areas, systems or items of plant below the main incoming meter. The purpose is not to create a wall of dashboards. It is to give engineering, operations and finance teams enough reliable information to separate productive energy use from avoidable waste, test the business case for improvements and confirm whether completed work has delivered the expected result.

Why the main meter is only the starting point

A supplier or fiscal meter is essential for billing and may also provide useful interval data. At whole-site level, that data can reveal broad changes in demand, overnight baseload and differences between working and non-working days. What it cannot usually show is which part of the process caused the change. A rise might come from an additional shift, a compressor fault, electric heating left enabled, a new production line or simply a more energy-intensive product mix.

AES's earlier article on how to identify which equipment consumes the most power introduced submetering as one way to find large consumers. The next step is to design the metering around the decisions the business actually needs to make. A meter installed without a clear question can produce accurate data that nobody uses; a smaller, carefully planned system can often be far more valuable.

Start with the questions the site needs to answer

Before selecting hardware, agree what the data should help the business understand. The questions may be operational, such as why one line uses more energy per batch than another, or commercial, such as whether a variable speed drive, motor replacement or controls change has a credible payback. Other sites may need to allocate energy costs between departments, compare shifts, identify non-production consumption or create a dependable baseline for an improvement programme.

Those questions determine the metering boundaries. If the decision concerns compressed-air performance, measuring only the main distribution board may be too broad. If the concern is the combined demand of a whole process area, installing a separate meter on every small motor may add cost and complexity without improving the decision. The right level of detail is the one that isolates a meaningful load and gives somebody a realistic opportunity to act on the result.

Design an industrial energy submetering hierarchy

Most useful systems follow the way energy and production are already organised. At the top is the incoming supply. Beneath it might sit transformers, main switchboards or major distribution boards, followed by production halls, lines, utilities and significant plant. The hierarchy does not have to reach every final circuit. It needs enough coverage for totals to make sense and enough separation to locate material changes.

In a food and drink factory, for example, the most useful boundaries might be processing, refrigeration, compressed air, packaging, cleaning systems and warehouse services. In an engineering business, the priority could be machining cells, extraction, heat treatment and central services. The names matter less than the relationship between the electrical distribution and the way the site manages production.

The design should also leave room for change. Spare communication capacity, consistent meter naming, accurate drawings and a clear approach to new connections make it easier to extend the system when a line is added or a process moves. Meter locations, current transformers, isolation arrangements and communications all require competent electrical design and installation. The appropriate equipment and method depend on the existing distribution, fault levels, environment, access constraints and the measurements required.

Match the data resolution to the process

A monthly total is too coarse for most investigations, but the finest possible interval is not automatically the best choice. A site with long, steady production runs may learn a great deal from relatively broad intervals. A fast-cycling compressor, batch oven or line with frequent stops may require much shorter intervals to make starts, idle periods and control behaviour visible. Higher-resolution data also creates more records to store, transfer and interpret, so the interval should be chosen to answer a defined question.

The UK Government's Energy Technology List criteria for automatic monitoring and targeting systems provide a useful reference point for system capability. They cover regular automatic collection, detection of missing data or communications failures, graphical reporting, data export, comparison between meters and exception reporting. Those are product eligibility criteria rather than a universal specification for every factory, but they underline an important principle: the value sits in dependable collection and usable analysis, not in the meter display alone.

Electricity consumption in kilowatt-hours is usually central, although some decisions also depend on demand, current, voltage, power factor or other electrical quantities. A wider energy plan may include gas, heat, steam or compressed-air flow. The measurement scope should be agreed by competent people who understand both the process and the electrical system, particularly where the findings may influence protection, capacity or investment decisions.

Put production context beside the energy data

A line that uses more energy this week may not be performing worse. It may have made more product, run a longer shift or handled a grade that needs more heating, mixing or cooling. Comparing raw kilowatt-hours without that context can reward lower output and disguise genuine inefficiency.

Useful comparisons often relate energy to an operational driver: kilowatt-hours per tonne, batch, finished unit or operating hour, for example. Depending on the process, the analysis may also need to recognise shift pattern, product mix, ambient temperature, cleaning cycles, planned downtime, rejects or changeovers. The aim is to choose a small number of measures that reflect how the factory actually works, not to collect every available tag from the control system.

This is consistent with the use of energy baselines and energy performance indicators in ISO 50001. A site does not have to pursue certification to benefit from the underlying discipline: define a meaningful baseline, relate energy to the factors that influence it and use the same basis when reviewing performance later.

Check the data before trusting the conclusion

A polished chart can still be built on poor data. During commissioning, meter labels and locations should be checked against current drawings, current-transformer ratios and direction should be verified, phases should be associated correctly, and clocks should be synchronised across the system. Communications failures and gaps need to be visible rather than silently filled, especially when a missing period could change the apparent baseline.