Energy Cost per Unit

Goal of the analysis:

Quantify the energy cost embedded in each good unit produced, understand the drivers (consumption, demand charges, time-of-use pricing, idle/startup losses), and reduce $/unit without compromising quality or service. Energy Cost per Unit links operations, maintenance, and finance: it translates OEE, changeovers, yield, and equipment efficiency into dollars, enabling executives to improve margin, prioritize efficiency investments (VFDs, compressed air programs, heat recovery), optimize schedules against tariffs, and align portfolio/pricing with cost-to-serve.

Data required:

• Utility billing and tariffs:
  • Electricity bills with line items: consumption (kWh), demand (kW) charges, time-of-use (TOU) rates, power factor penalties, riders/fees; interval data (15–60 min) if available.
  • Fuel/steam/gas consumption and tariffs (therms/MMBtu), compressed air valuation (if costed), water/sewer where material.
• Submetering and equipment telemetry:
  • Line/asset meters (kWh, kW, flow, pressure), PLC tags for speed/load, on/off states; compressor/boiler/oven meters.
  • Building loads (HVAC/lighting) if shared; battery/UPS where present.
• Production and operations context:
  • Good units by SKU/line/shift; runtimes, changeover/startup durations; OEE (Availability/Performance/Quality).
  • Routings and standard cycle times; lot sizes; product families; bottleneck identification.
• Quality and yield:
  • Scrap/rework counts and codes; First Pass Yield; startup scrap windows to quantify energy per good vs total.
• Environment and utilities systems:
  • Ambient temperature/humidity (HVAC load), compressed air pressures and leak survey results, steam trap survey outcomes.
• Financial and policy references:
  • Standard/target energy intensity (kWh/unit) if defined, capital project pipeline and hurdle rates, demand response participation rules.

Detailed step-by-step instruction on how to conduct the analysis:

1. Define metrics and scope.
   • Energy Cost per Unit ($/unit) = Total energy cost attributable to production ÷ Good units produced (period).
   • Energy Intensity (kWh/unit, MMBtu/unit) by energy type; Demand Cost share (%) = Demand charges ÷ Total electric cost.
   • Choose scope: plant/value stream and, where possible, line/SKU-level with submetering/allocation rules.
2. Assemble and align data.
   • Collect 12–16 weeks of utility bills and interval data; pull production and OEE by line/shift; extract submeter readings or estimate loads from nameplate × run hours.
   • Align to a common time base (15/30/60-minute intervals); normalize for shutdowns/holidays.
3. Separate base load from process load.
   • From interval profiles, identify non-production base load (nights/weekends) vs production-correlated load. Quantify base load $ and assign appropriately (e.g., area-hours) to avoid overstating per-unit variable energy.
4. Allocate shared utilities to lines/SKUs.
   • Direct: use line submeters where available.
   • Indirect: allocate by runtime-weighted standard power, demonstrated kW × hours, or regression of kWh vs throughput by line; allocate HVAC by area-hours or degree-days; allocate demand charges by contribution to coincident peaks (kW at peak interval).
5. Compute Energy Cost per Unit.
   • Per line/period: $ energy = Σ(kWh × TOU rate) + demand charges + fuel/steam cost + other riders.
   • $ per unit = $ energy ÷ good units; also compute kWh/unit and split by state: run, idle, setup/startup (using state tags or approximations).
   • Adjust for yield: report both “per produced unit” and “per good unit” to expose scrap/rework penalty.
6. Decompose drivers.
   • Cost bridge: Base load → Run consumption → Idle/standby → Setup/startup → Demand charges → Power factor penalties.
   • TOU bridge: On-peak vs off-peak $ and kWh shares; identify load during peak intervals that drives demand.
7. Segment and compare.
   • By line, SKU family, shift, utilization bands, and changeover intensity; highlight bottom quartile kWh/unit and $/unit cells.
   • Plot kWh/unit vs utilization to estimate fixed vs variable components (intercept ≈ fixed kWh/hour); plot $/unit vs FPY/changeover/downtime.
8. Identify waste and improvement targets.
   • Quantify idle/standby energy share; startup spikes linked to changeovers; compressed air leaks (baseline via compressor load during non-production).
   • Flag lines with high on-peak share or frequent coincident starts causing demand peaks.
9. Scenario and what‑if analysis.
   • Model impacts: SMED (startup time −30–50%), shift to off-peak (reroute waves, add late pickup), VFDs on fans/pumps (−15–30% kWh), compressed air leak repair/pressure reduction (−10–20%), oven/burner tuning and insulation (−5–15%), heat recovery (process dependent), power factor correction (reduce penalties/demand).
   • Estimate $/unit reduction, demand charge reduction, payback/IRR; include M&V plan (pre/post baseline, weather/volume normalization).
10. Integrity checks.
    • Reconcile total allocated kWh to utility totals (±1–2%); ensure “good units” match MES/shipping; avoid double-counting demand charges.
    • Validate allocation assumptions with spot metering; keep “unknown/other” <10% of cost.

Format of the output of analysis:

• Executive scorecard: Energy Cost per Unit ($/unit) and Energy Intensity (kWh/unit) by plant/line/SKU family; Demand cost share %; on- vs off-peak mix; trend vs target.
• Heatmaps: $/unit and kWh/unit by line × shift and by SKU family; idle/setup share by line.
• Load profiles: interval kW with production overlays; peak events and coincident start-ups; TOU cost breakdown.
• Cost bridges: base/run/idle/setup/demand components; yield penalty ($/good unit due to scrap/rework).
• Driver charts: kWh/unit vs utilization; $/unit vs FPY, changeover minutes, downtime.
• Scenario/ROI panel: projected $/unit reduction, demand charge savings, capex, payback, and M&V plan.

How to interpret results:

• High $/unit with low utilization: Fixed/idle energy dominates; improve scheduling (campaigns), auto-standby/shutdown, consolidate runs to dilute base load.
• High $/unit on changeover-heavy lines: Startup/heat-up energy and scrap inflate cost; run SMED, centerline recipes, and pre-heat/preset off-line.
• Large demand charge share (≥25–40% of electric bill): Peak coincident starts; stagger equipment startups, use soft-start/VFDs, shift load off-peak; consider demand response.
• High compressed air or HVAC share: Leak/pressure or comfort/process setpoint opportunities; pressure down 1 bar ≈ ~7% energy reduction (directional); optimize HVAC setpoints and sequencing.
• $/unit high despite good kWh/unit: Production skewed to on-peak tariffs; reschedule where feasible or renegotiate tariffs.
• Energy penalty from poor yield: Low FPY means more kWh per good unit; address quality first for compounding benefit.

Steps a company can take to improve on this measure:

• Operational and scheduling levers:
  • Level releases (heijunka), campaign by family, enforce frozen windows; stagger startups to avoid peaks; add late carrier pickups to enable off-peak production.
  • SMED to cut setup/startup time; centerline parameters and first-article checks to reduce warm-up scrap.
• Equipment and controls:
  • Install VFDs on variable-torque loads (fans/pumps); upgrade motors; tune burners/ovens and improve insulation; add auto-standby and shutdown logic.
  • Implement power factor correction if penalties apply; consider heat recovery (e.g., oven exhaust to pre-heat air/water) where feasible.
• Utilities system optimization:
  • Compressed air: leak detection/repair program, lower pressure setpoint, proper storage/controls; right-size compressors.
  • Steam/HVAC: trap surveys, condensate return, insulation audits, economizers; optimize HVAC schedules and setpoints.
• Quality and reliability:
  • Raise FPY; reduce rework/scrap; deploy TPM/CBM to cut downtime (avoids energy spent on restarts and idling).
• Tariff and procurement:
  • Enroll in optimal TOU/demand tariffs; evaluate aggregation and demand response incentives; negotiate fuel/steam rates where applicable.
• Data and governance:
  • Expand submetering on top value streams; integrate energy KPIs (kWh/unit, $/unit) into daily tier boards; validate savings with M&V.
• Example scenarios:
  • If $/unit is elevated by demand charges (35% of electric cost), stagger line startups by 5–10 minutes and add soft-start/VFDs; expect 10–20% demand reduction and −$0.03–$0.06/unit (context-dependent).
  • If compressed air baseline shows 25% load during non-production, execute a leak program and lower pressure 0.2–0.3 bar; target 10–20% compressor kWh reduction and −$0.02–$0.04/unit.
  • If setup/startup energy is 12% of line consumption, run SMED and pre-heat change parts; expect startup kWh −30–50% and −$0.01–$0.03/unit.

Benchmark comparisons:

General benchmarks (directional):

• Energy as % of COGS: often 1–5% in discrete assembly; 5–20%+ in energy‑intensive processes (ovens, kilns, foundry, chemicals).
• Demand charges can be 20–50% of electric bills under TOU tariffs; compressed air commonly 10–30% of plant electricity where prevalent.
• Typical savings potentials: VFDs on fans/pumps 15–30% kWh; compressed air leak/pressure programs 10–20%; oven/burner tuning and insulation 5–15%; scheduling/off‑peak shifts reduce $/unit where tariffs are steep.

Constructing internal benchmarks:

• Track weekly $/unit and kWh/unit by line/SKU family and shift; publish quartiles and set alerts (e.g., $/unit >1.5× median or idle share >15%).
• Benchmark lines with similar products: target top‑quartile kWh/unit as internal standard; set idle energy <10–15% of total line kWh.
• Monitor demand events (top 10 peak intervals) and on‑peak share; aim to shift ≥10–20% of flexible load off‑peak.
• Rebaseline after SMED/TPM, equipment upgrades, or tariff changes; maintain an M&V registry to lock in verified savings.