Measurement & KPI Guide — know whether changes actually cut energy and emissions

Good actions need good verification. This guide explains what to measure, how to normalize for activity, and which KPIs help operations make reliable decisions.

What to measure

• Whole-site electricity and fuel: Useful for overall tracking and verifying large interventions.
• Submetered systems: HVAC, compressed air, process lines, boilers—measure at the system or major equipment level where possible.
• Process activity: Throughput, production hours, beds occupied, patient visits, or service throughput—metrics you can use to normalize energy use.
• Ambient and process variables: Temperatures, pressures, run hours, setpoints—these help explain why energy changed.

Simple, robust KPIs

• Energy intensity: Energy per unit of activity (kWh per unit produced, kWh per bed-day, kWh per square foot per hour).
• Absolute consumption: Daily or weekly kWh for the system under evaluation.
• Runtime and load factor: Percent of time equipment runs and average kW while running.

Normalizing for activity

Always compare similar operating conditions. If production volume, occupancy, or outdoor temperature changed between test periods, normalize the energy by the appropriate activity metric or include the operating condition in a simple regression for the pilot period.

Verification and persistence

Collect at least one full comparable cycle (shift, day, or week depending on the process) before and after a change. For seasonal systems, consider a longer run to ensure persistence. Record the measurement period, data source, any adjustments made, and the person responsible for the validation.

Using dashboards and huddles

Make verified results visible in your Sustainability Metrics Huddle so teams can learn from experiments. Use SCADA or submeters to automate basic alerts (e.g., changes in run-time or sudden increases in kW) but tune alerts to avoid fatigue.

Short checklist for a credible measurement

• Define baseline period and activity normalization method.
• Capture pre-change readings from the same meters and same time-of-day.
• Record any confounding events (maintenance, production changes, outages).
• Collect post-change data for a comparable period and compute normalized KPI changes.
• Document and present findings in the next huddle with a recommended action (scale, tweak, rollback).

This guide is intentionally practical: you don’t need perfect metering to learn. Start with the best available data, document uncertainty, and improve measurement fidelity over time using the Facilities SCADA Cluster and metering recommendations found elsewhere in the hive.