Energy Metering Plan and KPI Definitions

Purpose and quick orientation

This guide helps operations teams choose meter locations, sampling rates, aggregation strategy, and a small set of actionable KPIs so you can spot energy hotspots, measure improvements, and prove savings. It focuses on pragmatic, low‑cost instrumentation and a pilot approach: baseline one production line or process, capture data, fix the easiest problems, measure results, then scale.

Plan overview (at a glance)

1. Map processes and energy flows.
2. Select meter points (main, subpanels, critical machines, process loads, HVAC, compressors).
3. Choose meter types & sampling frequency.
4. Decide on data aggregator & storage.
5. Define baseline period and normalization rules.
6. Pick a small KPI set and reporting cadence.
7. Run a pilot: baseline, quick fixes (30 days), track savings, then scale.

1. Map processes and energy flows

Sketch the production line, utilities, and building systems. Identify where electricity enters the site, where it splits to panels and process equipment, and where major loads live (compressors, ovens, chillers, motors, pumps, air dryers). A simple floor map with numbered candidate meter points is enough for a pilot.

2. Select meter points (practical guidance)

• Site main/utility meter — captures total energy; useful for validating aggregation and cost allocation.
• Feeder/subpanel meters — group related loads (one line, one product family, or one building zone).
• Critical machines/processes — motors, ovens, presses, compressors where control or behavior changes can yield savings.
• Ancillary systems — air compressors, HVAC, pumps, chillers, lighting circuits that often produce quick wins.
• Portable/logging options — clamp meters or smart plugs for short trials before committing to permanent CTs.

Practical rule: start with one production line or one major utility group where you can directly link energy to output or run‑hours.

3. Meter types and sampling frequency

Match meter resolution to the problem:

• High resolution (1s–1min) — use for variable-speed drives, demand spikes, motor start analysis, or transient events.
• Medium resolution (1–5 min) — good for production line energy per unit, compressed air usage patterns, shift comparisons.
• Low resolution (15 min–hourly) — acceptable for long‑term trend analysis and cost allocation but may mask short events.

Choose CTs (current transformers) sized for the feeder and consider vendor meters that provide kWh, power (kW), pf, voltage, and time stamps. Ensure clocks are synchronized or timestamped in UTC to avoid misaligned comparisons.

4. Data aggregator and storage

Decide where to collect and store time series data. Options include local dataloggers, an on‑prem gateway, or a cloud aggregator. Key needs:

• Time‑aligned series with unit and production counts.
• Reliable ingestion and retention for the baseline period (30–90 days) and ongoing tracking.
• Exportable CSV or API access for verification and analysis.

5. Baseline and normalization

Define a baseline that matches typical operations: 30–90 days is common. Normalize energy to production volume, run‑hours, and weather (degree days) where relevant. Document what is excluded (planned downtime, outages, maintenance) and keep a change log of process or product mix changes that could affect energy per unit.

6. KPI definitions (clear, actionable formulas)

• kWh per unit = (kWh measured for the product line during period) / (units produced during same period). Use same timestamps for both numerator and denominator.
• Energy cost per SKU = kWh per SKU × average $/kWh (include demand charges allocation if needed).
• Peak demand by shift = maximum kW observed during the shift. Report with time and contributing panels if available.
• Load factor = average kW / peak kW over the evaluation period. Low load factor often signals opportunities to reduce peaks.
• kW per operator or per run‑hour — useful where output is difficult to count.

Always state the measurement window (e.g., daily shift, weekly) and the production normalization used.

7. Pilot plan (30–60 day example)

1. Install meters on one line/subpanel and the site main (week 1).
2. Collect baseline data for 14–30 days, capturing production and operating modes (weeks 2–4).
3. Identify 3–5 quick wins (scheduling, setpoints, compressed air leaks, lighting, idling) and implement low‑cost fixes (week 5–6).
4. Measure post‑fix performance for 14–30 days and calculate verified savings vs baseline, normalized for production (weeks 7–8).
5. Document lessons, ROI, and a rollout plan for additional meters/sites.

Assign a single owner for the pilot who is accountable for data integrity, production counts, and following up on fixes.

8. Measurement & verification (practical approach)

For operational fixes, a simple M&V approach is sufficient:

• Baseline → implement change → post period with same normalization
• Adjust for production and major weather differences
• Report gross kWh saved and convert to $ using actual tariff including demand impacts if measurable
• Keep evidence: meter screenshots/exports, production logs, timestamps of implemented controls

9. Data quality checks and common pitfalls

• Clock mismatch between meters and production system—sync before baseline.
• Meters placed upstream of variable groups—can hide machine‑level behavior. Use subpanel or machine meters where needed.
• Insufficient sampling masks transients—raise resolution if you need to analyze starts/stops.
• For multi‑product lines, failure to normalize by SKU mix produces misleading KPIs.
• One‑off manual changes during the baseline period can bias results—record exceptions.

10. Quick wins checklist

• Fix compressed air leaks and add automatic shutdown for idle periods.
• Reduce lighting during non‑production hours or install vacancy sensors.
• Adjust setpoints and schedules for HVAC and steam systems to match occupancy/shift.
• Turn off unused motors, conveyors, or ovens between batches where safe.
• Implement simple motor VFD controls or soft starts where starts are frequent and spikes matter.

Next steps

Start with a single line pilot, capture 30 days of aligned energy and production data, implement low‑cost fixes, and verify savings. Use the KPI formulas above and document normalization rules. If the pilot shows measurable savings and reliable data, scale metering and formalize an ongoing monitoring and ownership model.

Note: This guide is operationally focused and does not replace certified energy engineering, permitting, or detailed capital project design.