Home › Technical articles

Đọc bằng tiếng Việt

Power monitoring systems (PMS) in a factory: energy meters, demand, power quality and the data path to SCADA

TDL Industrial Automation · Last updated:

From Rockwell's PowerMonitor 5000 and PowerMonitor 1000 manuals and AVEVA's documentation: what a meter records, how demand and power factor are calculated, meters at the utility feed, feeders and machines, logs in the meter, EtherNet/IP and Modbus to SCADA, energy counters and reports.

A power monitoring system (PMS) reads the power meters on a plant's electrical distribution and records energy, demand, power factor and power quality events, so engineers can see where the electricity goes and what happened when the supply misbehaved. It is made of meters on the incomers, feeders, MCCs and utility equipment, a network to read them, data servers, and a SCADA with a historian that turns the readings into alarms and energy reports. How a PMS relates to an FMCS, a BMS and an MES is covered in What is FMCS; this article follows Rockwell Automation's PowerMonitor documents and AVEVA's documentation along the path from meter to report.

1. What a power meter records

Rockwell's PowerMonitor 5000 (Bulletin 1426) manual lists what the meter measures: voltage line to line and line to neutral, current, frequency, voltage and current unbalance, real, reactive and apparent power, true and displacement power factor, energy in kWh, kVARh and kVAh, kW, kVAR and kVA demand, projected demand, demand power factor and crest factor (1426-UM001L-EN-P pp. 20–21). Rockwell's product page adds that it samples four voltage and four current channels every cycle.

The PowerMonitor 1000 (Bulletin 1408) manual explains the main quantities (1408-UM002E-EN-P pp. 61–65):

  • Energy is kept as forward, reverse and net totals. When the kWh total rolls over to zero, the GWh total goes up by one.
  • Demand is the average power over a demand period, 15 minutes by default. Setting more than one period gives a sliding-window demand. The end of each interval comes from the internal timer, status input 2 (an external end-of-interval contact), a controller command or an Ethernet demand broadcast.
  • Projected demand is a linear projection to the end of the current interval; demand power factor is kW demand divided by kVA demand.
  • True power factor is reported per phase and in total, lagging as negative and leading as positive.

Power quality is where the PowerMonitor 5000 models differ. The M6 adds harmonics up to the 63rd order, waveform capture and power quality event classification; the M8 adds flicker, sub-cycle transient capture, harmonics up to the 127th order and interharmonics (p. 18). Sag and swell detection is listed for all models, EN 50160 weekly and yearly logs for the M8 only (p. 21).

Measurement Where it is measured What it is used for
Energy – kWh, kVARh, kVAh main feeds, feeders, process lines, machines consumption reporting, cost allocation by actual use, shadow billing
Demand – kW, kVA, projected utility feed, plant areas demand analysis, load profiling, keeping demand under a peak limit
Power factor utility feed, feeders power factor correction to cut utility penalties
Voltage, current, unbalance feeders, MCCs, distribution panels a central view of the distribution for trending, alarms and load shedding
Harmonics, sags, swells utility feed, where power enters power quality monitoring to protect equipment and reduce downtime
Status inputs – WAGES PowerMonitor 5000 inputs water, air, gas, electricity and steam data

The table follows Rockwell's PowerMonitor Products Application Guide (1400-AT001A-EN-P pp. 5–10, 15, 30).

2. Where the meters sit

The application guide starts with "metering the envelope": energy monitoring on the plant main feeds, which shows total consumption, lets the utility bill be checked with a shadow bill and points to the largest users. Submeters then go on process lines and utility equipment such as compressors and boilers, and their energy is correlated with production indicators (p. 5). For each level it names a meter (p. 15):

  • Utility feed: a PowerMonitor 5000, to catch power quality problems where power enters the plant.
  • Feeders, MCCs, control panels: a PowerMonitor 1000, most often for submetering at cell or site level. Its manual states that it is not a circuit protective device (1408-UM002E-EN-P p. 7).
  • Machines: a PowerMonitor 500, which also covers motors on across-the-line starters that give no energy data.

The current PowerMonitor 1000 models are BC3 (basic consumption), TS3 (adds voltage, current, frequency and unbalance) and EM3 (adds demand and the load factor log); TR1, TR2, EM1 and EM2 are discontinued (1408-UM002E-EN-P pp. 8, 10).

3. What the meter keeps by itself

The PowerMonitor 5000 keeps energy, data, min/max, load factor, time-of-use, event, setpoint and alarm logs on every model, plus power quality, waveform, trigger and snapshot logs on the M6 and M8. It has setpoints, three relay outputs, one KYZ pulse output and four status inputs (1426-UM001L-EN-P p. 21), and the product page lists configurable alarms for up to 20 events. The PowerMonitor 1000 keeps time-of-use, energy, min/max and status logs, with the load factor log on the EM3 (1408-UM002E-EN-P p. 10). Both serve web pages for commissioning and wiring diagnostics, from which data logs can be exported (1400-AT001A-EN-P p. 18).

4. Communication: from the meter to the SCADA

The PowerMonitor 5000 has a native EtherNet/IP port with a web page, and optionally DeviceNet or ControlNet (1426-UM001L-EN-P pp. 17, 19). The PowerMonitor 1000 has RS-485 on every model, with DF1, Modbus RTU and DH-485, and on -ENT models an Ethernet port with EtherNet/IP, Modbus TCP and a web page, at 10 Mbps half duplex (1408-UM002E-EN-P pp. 8–12).

Power monitoring: from meters to energy reportsAt the top, three meter locations: A, the utility feed with a PowerMonitor 5000 for energy, demand and power quality; B, feeders and MCCs with a PowerMonitor 1000 for energy, demand and power factor; C, utilities and machines with submeters and a PowerMonitor 500. The meters connect over EtherNet/IP, Modbus TCP and Modbus RTU on RS-485. Data takes one of three paths: a Logix controller with the Add-on Profile, Modbus TCP direct, or a Modbus gateway for RS-485 meters. Data servers with the ABCIP and MBTCP drivers pass it to the SCADA and historian, which produce the energy reports. Based on Rockwell Automation and AVEVA documentation.Power monitoring: from meters to energy reportsA · Utility feedPowerMonitor 5000energy, demand, harmonics,sags and swellsB · Feeders and MCCsPowerMonitor 1000energy, demand,power factorC · Utilities and machinessubmeters,PowerMonitor 500compressors, boilers, motorsNetwork: EtherNet/IP · Modbus TCP · Modbus RTU on RS-485Logix controllerAdd-on Profile, meter dataas tagsModbus TCP directmeters with an Ethernet portModbus gatewayRS-485 meters, oneunit ID eachData serversABCIP driver to the controller, MBTCP driver for Modbus TCPSCADA and historianalarms, trends, counter retrieval of energy totalsEnergy reportsconsumption by area and shift, demand, power factor, power quality events, cost allocationBased on Rockwell Automation 1400-AT001A, 1426-UM001L, 1408-UM002E and AVEVA CommunicationDrivers 2026, Counter retrieval (read 15 November 2026). Figure: drawn by TDL.
Figure 1: Power monitoring: from meters to energy reports. On a phone, swipe sideways to see the whole figure.

Through a Logix controller. Add-on Profiles in Studio 5000 Logix Designer bring meter data into the controller as tags: 16 parameters chosen by number on the PowerMonitor 1000, full configuration and monitoring on the PowerMonitor 5000 (1400-AT001A-EN-P pp. 19, 22, 24). The controller can combine them with breaker and switch signals, and the SCADA reads the controller as usual; on AVEVA through the ABCIP driver, see AVEVA OI Server ABCIP.

Directly over Modbus TCP. AVEVA's MBTCP Communication Driver (Communication Drivers 2026) has a generic ModbusPLC connection for Modbus controllers that are not on its supported list but follow the Modbus specification. Its settings show what must match the meter's register map: port 502 by default, a reply timeout of 3 seconds by default, one-based addressing unless zero-based is selected, byte swapping, register order for 64-bit values and block sizes. Serial devices behind a Modbus TCP bridge go under a ModbusBridge object as ModbusPLCRS connections, each with its own unit ID.

Time. Events from different meters only line up if their clocks agree. The PowerMonitor 1000 sets its clock by SNTP (1408-UM002E-EN-P p. 60), and the PowerMonitor 5000 supports network time synchronisation, IEEE 1588 and network demand synchronisation (1426-UM001L-EN-P p. 21).

5. From history to energy reports

Energy totals are counters, so a report needs the increase over each hour or shift. AVEVA's counter retrieval mode returns that increase for each cycle, including rollovers, using a rollover value set for each tag in the Historian, and it works with floating-point counters (Operations Management Interface 2026). How a historian stores data: what is a process historian.

Rockwell's guide lists what the data is used for: consumption reporting to catch leaks and waste, demand analysis, submetering, cost allocation by actual use rather than floor area, distribution monitoring, power quality, load profiling and shadow billing (1400-AT001A-EN-P pp. 6–7). It also describes control on the same data: demand management projects demand so a peak limit is not exceeded, load management sheds loads in a set order, and power factor correction capacitors reduce utility penalties (p. 30).

Two Rockwell energy software products appear in these sources. The 2018 guide describes FactoryTalk EnergyMetrix as a central energy database used through a web browser, with consumption, demand, billing and cost allocation reports and a power quality report of the sags and swells PowerMonitor units detect (pp. 26–29). FactoryTalk Energy Manager is a cloud software-as-a-service application on FactoryTalk DataMosaix, with dashboards for energy consumption, intensity, costs and emissions, energy baselines and measurement and verification. Meter alarms need priorities like any plant alarm: see ISA-18.2.

What we check first

This is TDL's practice, not a Rockwell Automation or AVEVA requirement.

  • We list every meter from the single-line diagram: the incomer, feeder or MCC it sits on, its model, CT and PT ratios and port.
  • We check that each model gives the values the report needs; on a PowerMonitor 1000, demand needs the EM3.
  • We choose the path per meter: EtherNet/IP into a controller through the Add-on Profile, or Modbus into the SCADA driver, with gateway and unit IDs for serial meters.
  • For Modbus, we set addressing base, byte swap and register order from the register map, then compare one known value with the meter's web page before building the rest.
  • We confirm the demand period and its synchronisation source, and one time source for meters and servers.
  • We set a rollover value for every energy tag and check one report period by hand against the meter's energy log.
  • We keep the meters on the control network, in their own zone: see IEC 62443.

How TDL approaches it

At Samsung Electronics HCMC CE Complex (SEHC), TDL maintains and upgrades the FMCS and the power monitoring system under a yearly maintenance contract: 17 servers, 14 workstations and 37 PLCs on ControlLogix and AVEVA System Platform, with Modbus TCP and EtherNet/IP. The scope includes preventive maintenance with backups and restore tests, ControlLogix firmware migration and redundancy reconfiguration, extension of the AVEVA System Platform SCADA, Modbus TCP to EtherNet/IP communication, and FAT and site commissioning (project page). TDL is a System Integrator in the Rockwell Automation PartnerNetwork and a System Integrator member of the AVEVA Partner Ecosystem. Services: facility monitoring (FMCS) and SCADA and historian.

Frequently asked questions

What is the difference between a PMS and an FMCS?

A PMS watches the electrical network: energy, demand, power factor and power quality from meters on incomers, feeders and machines. The FMCS article explains how the two relate; at SEHC, TDL maintains the FMCS and the PMS together on ControlLogix and AVEVA System Platform.

Where should power meters be installed in a factory?

Rockwell's application guide starts with the plant main feeds, then submeters on feeders, process lines and utility equipment such as compressors and boilers. It puts a PowerMonitor 5000 at the utility feed, a PowerMonitor 1000 on feeders and MCCs and a PowerMonitor 500 at machine level.

How is electrical demand calculated?

Demand is the average power over a demand period. On a PowerMonitor 1000 the period is 15 minutes by default, more than one period gives a sliding window, and the end of each interval can come from the internal timer, status input 2, a controller command or an Ethernet broadcast.

Can a PowerMonitor be read over Modbus?

The PowerMonitor 1000 supports Modbus RTU on its RS-485 port and Modbus TCP on its Ethernet port, as well as EtherNet/IP. The PowerMonitor 5000 manual lists EtherNet/IP as its native network, with optional DeviceNet or ControlNet.

How is a meter's energy total turned into consumption per shift?

The energy total is a counter that rolls over, so a report needs the increase per period, not the raw value. AVEVA's counter retrieval mode returns that increase for each cycle, including rollovers, using a rollover value set for each tag in the Historian.

Sources

Read 15 November 2026. This article paraphrases the sources and does not quote them; read the original documents before designing or working on a system.

Related services:
SCADA & historian Facility and utility monitoring (FMCS)