Energy efficiency is the most reliable “first ton” in decarbonization. It reduces emissions immediately, lowers operating cost, and improves resilience. It also makes every other decarbonization move easier: smaller loads reduce the scale of renewable procurement, electrification equipment, storage, and grid upgrades you must fund. Despite this, many organizations treat efficiency as a set of projects, owned by no one, measured inconsistently, and allowed to erode after the first wave of activity.
This chapter is a practical guide to making efficiency repeatable. We start with where efficiency typically lives in the P&L and the asset base, then clarify when it should be the first lever. We outline the minimum data required to make efficiency decision-grade, provide a step-by-step approach to designing and executing an efficiency program, and close with how to embed continuous improvement so savings persist beyond the first wave.
7.1 Where Efficiency Typically Lives in the P&L and Asset Base
Efficiency is rarely “missing.” It is usually spread across budgets, hidden in allocations, and diluted by organizational seams. The first task is to see efficiency as a system of costs and constraints rather than as an engineering hobby. When leaders do not see the full economics, efficiency gets undervalued and underfunded.
Energy cost stack: The full set of costs associated with energy use, including commodity charges, demand charges, network fees, fuel costs, maintenance impacts, and productivity losses that energy waste enables.
On the P&L, energy shows up beyond the utilities line. Demand charges and peak-related tariffs can be a material share of electricity spend, yet they are often treated as uncontrollable overhead. Fuel for fleets is managed by logistics, not facilities. Steam and compressed air costs may never appear explicitly; they are embedded in fuel and maintenance budgets. Meanwhile, energy waste creates second-order costs that do not look like “energy” at all: scrap, rework, unplanned downtime from overheated equipment, short cycling that accelerates component wear, and throughput loss from unstable processes.
A useful habit is to translate “energy waste” into the language of the business. For example, a steam leak is not only a fuel cost; it is also a boiler cycling issue that increases maintenance, a condensate quality issue that can reduce process stability, and sometimes a safety risk. Likewise, poor HVAC scheduling is not only a kWh issue; it can affect comfort, productivity, and tenant complaints. When you capture these linkages explicitly, efficiency stops being a discretionary expense and becomes an operating improvement with financial and reliability outcomes.
Efficiency also shows up as a constraint, not just a cost. Many sites are limited by electrical service, transformers, steam header capacity, cooling capacity, or compressor capacity. When you reduce demand, you may avoid or defer capacity upgrades and create headroom for electrification.
In the asset base, efficiency concentrates in a small number of systems that repeat across industries. Physics is predictable: conversion, distribution, and end use. The same “big rocks” show up in factories, data centers, hospitals, retail networks, and campuses.
Conversion systems: Boilers, furnaces, ovens, chillers, refrigeration racks, power supplies, and onsite generation.
Distribution systems: Steam networks, hot water loops, compressed air headers, chilled water loops, ductwork, and electrical distribution with power quality issues.
End-use systems: Motors, pumps, fans, conveyors, process heating, HVAC terminal units, lighting, and IT loads.
Loss map: A structured view of where energy is wasted, typically broken into avoidable demand, conversion losses, distribution losses, and end-use inefficiencies.
A loss map helps because it turns “energy” into causes that operators recognize. Avoidable demand includes running equipment when not needed, redundant standby loads that became permanent, and setpoints that drift upward “just in case.” Conversion losses include poor boiler tuning, low part-load chiller performance, and refrigeration systems operating outside design conditions. Distribution losses include steam leaks, failed steam traps, compressed air leaks, poor insulation, and pressure setpoints that are higher than necessary. End-use inefficiencies include unstable controls, simultaneous heating and cooling, and equipment that is oversized or poorly matched to duty cycles.
For most organizations, the highest-value way to segment demand is by behavior over time. Base load: energy that is present even when production or occupancy is low. Variable load: energy that tracks throughput, weather, or utilization. Peak load: short periods of high demand that drive demand charges and capacity constraints. Base-load reductions deliver steady savings. Variable-load reductions improve intensity. Peak reductions can unlock disproportionate cost savings and avoid costly upgrades.
Finally, do not underestimate operational discipline. Many efficiency opportunities are not technology gaps; they are governance gaps. Controls are overridden, maintenance is deferred, and setpoints are adjusted without accountability. The difference between average and top-quartile energy performance is often the presence of routines: clear ownership, reliable data, and standard work that prevents drift and makes “energy waste” visible the same way quality defects become visible.
7.2 When Energy Efficiency Is the Right First Lever
Efficiency should usually be the first lever when you need near-term reductions, when the organization’s data foundation is weak, or when deeper decarbonization options are constrained by infrastructure and capital cycles. “Efficiency first” is a sequencing choice, not a philosophy that excludes electrification, renewables, or process change.
Efficiency-first decision: A deliberate sequencing choice to prioritize demand reduction and operational efficiency before scaling supply-side decarbonization, to reduce load, cost, and dependency risk.
Efficiency is the right first lever in five common situations. First, when energy is a material operating cost and margins are under pressure; efficiency becomes self-funding abatement with short payback. Second, when you have near-term commitments (often 12–36 months) and cannot rely on long-lead projects like major retrofits, interconnections, or new fuel supply. Third, when electrification and renewables face constraints: limited grid capacity, long interconnection timelines, uncertain permitting, limited contractor capacity, or sites that cannot tolerate extended downtime.
Fourth, efficiency is the right first lever when reliability is a priority. Many measures reduce failure risk: better cooling management, improved steam and condensate integrity, reduced compressor cycling, and optimized equipment sequencing that avoids short cycling and thermal stress. These benefits matter in sectors where downtime costs dwarf energy savings. Fifth, efficiency is the right first lever when you need to build organizational capability. A well-run efficiency program creates the measurement routines, governance, and delivery cadence that later enable larger decarbonization programs to scale without collapsing under verification and claims scrutiny.
There are also cases where efficiency is not the limiting factor. In a Scope 3–dominated company, facility efficiency is necessary but insufficient; it should not consume leadership bandwidth that must be spent on procurement, product design, and supplier engagement. In a utility or upstream energy producer, the core emissions pathway may depend on portfolio transition and methane control; facility efficiency still matters, but it is not the main driver. The practical answer is to run efficiency as a disciplined program in parallel, while directing the majority of strategic attention to the levers that dominate the footprint.
The most common failure mode is treating efficiency as a “projects list” without a role in the broader pathway. In a strong decarbonization portfolio, efficiency plays three roles. Immediate abatement: deliver measurable reductions quickly. Cost and capacity release: lower peak and base load so electrification and renewables are smaller and cheaper. Capability building: establish the data, routines, and ownership model that makes later measures verifiable and durable.
7.3 Data Required: Asset Inventories, Load Profiles, Operational Data
Efficiency programs do not fail because teams lack ideas; they fail because teams cannot prove impact and therefore cannot sustain funding and trust. The goal is not perfect data. The goal is a minimum viable dataset that supports prioritization and credible verification, with a plan to improve data quality where savings and scrutiny are greatest.
Minimum viable energy dataset: The smallest set of asset, consumption, and operational data that enables a team to identify efficiency opportunities, build business cases, and verify savings with reasonable confidence.
Start with an asset inventory that is operationally meaningful. For each major system, capture: location, served area or process, capacity, age, efficiency class if known, control method, and maintenance status. For industrial sites, include major process units and the utilities systems that feed them. For buildings portfolios, include HVAC plant, air distribution, refrigeration where relevant, lighting, and controls. For fleets, capture vehicle types, duty cycles, mileage, fuel type, and idle time.
Next, build load profiles. For electricity, interval data is the single most useful input because it reveals base load, peaks, and operational scheduling issues. For fuels, daily consumption aligned to operating hours is often enough to start. For steam and compressed air, production of utility (boiler output, compressor kW) paired with pressure and flow indicators can reveal leaks and control issues. The key is to segment base, variable, and peak behavior; that segmentation drives both the economics and the choice of interventions.
Operational driver data makes verification credible. Capture the minimum set of drivers that explain energy variation: production volume, operating hours, weather for HVAC-driven loads, and major operational events (shutdowns, product mix shifts, maintenance outages). Without driver data, teams argue about whether savings are real. With driver data, teams can normalize and separate operational change from demand change.
Finally, define verification rules upfront. Agree on baseline period, normalization method, and sign-off process before implementation begins. If verification is designed after the fact, it becomes a negotiation rather than a control.
Baseline period: The historical period used to define expected energy use absent a measure.
Normalization: Adjusting energy data for production, operating hours, and weather to enable fair comparison.
Verification: A repeatable method to confirm savings, using meters where possible and transparent estimates where not.
7.4 Step-by-Step: Designing and Executing an Energy Efficiency Program
An efficiency program should be managed like any other performance program: clear goals, a pipeline, disciplined execution, and controls that prevent backsliding. The steps below are designed to work at a single site or across an enterprise.
Step 1: Define program goals and the “use case” for efficiency: Be explicit about whether the program is primarily targeting cost reduction, emissions reduction, capacity release for electrification, or reliability. This matters because it changes prioritization. A cost-led program will value demand-charge reduction. An emissions-led program will value kWh and fuel reduction. A capacity-led program will value peak shaving and base-load reduction in constrained locations.
Step 2: Assign ownership and decision rights: Name a program owner with authority to coordinate operations, maintenance, engineering, procurement, and finance. Create site owners who can change operating practices. Clarify who can change setpoints, schedules, sequences, and maintenance priorities, and how conflicts are resolved when production stability is at stake.
Step 3: Establish baseline and segmentation: Build a baseline by site and major end use, then segment into base, variable, and peak. Identify top consumers and top cost drivers. If you have many locations, start wide and then go deep: a high-level baseline for all, then detailed analysis for the sites that represent most energy, cost, and risk.
Step 4: Build the measures pipeline: Combine analytics, walk-throughs, and operator input. Analytics finds anomalies and peak drivers. Walk-throughs find leaks, overrides, and scheduling errors. Operators identify workarounds and unstable controls that drive waste. Convert ideas into measures with a clear mechanism, owner, and implementable unit (a site, a system, or a line). Avoid measures that are so broad that no one can implement them.
Most measures fall into five repeatable buckets.
Operational tuning: Setpoints, schedules, equipment sequencing, and control loop stability.
Maintenance and loss reduction: Leak repair, steam trap programs, insulation restoration, and refrigerant management.
Equipment upgrades: Drives, motors, boilers, chillers, refrigeration packages, and lighting.
System redesign: Heat recovery, right-sizing, network redesign, and elimination of redundant loads.
Peak management: Load shifting, demand response, and thermal storage where economics justify it.
Step 5: Create a “measure card” for each initiative: A measure card is a one-page structured description that makes measures comparable and executable. It keeps the pipeline from becoming a collection of vague ideas and helps finance and operations review quickly.
Measure definition: What changes, where, and why it reduces energy.
Boundary: Which meters or bills will reflect the change, and which loads are included.
Savings logic: The formula or method that converts activity change into kWh, fuel, and CO2e reduction.
Cost and funding: Incremental capex, opex, expected savings, and payback range.
Feasibility: Downtime needed, safety and quality constraints, permitting, and supplier lead times.
Owner and timing: Implementation owner, start date, completion date, and verification date.
Risks and mitigations: What could cause underperformance and how it will be prevented.
Step 6: Quantify impact and economics consistently: For each measure, estimate energy savings, emissions savings, capex, opex, and operational impacts using one consistent method. Use ranges and document drivers of uncertainty. Separate engineering estimates from verified savings. Apply overlap rules so measures do not claim the same tons twice, especially when efficiency reduces the load that renewables or electrification would otherwise address.
Step 7: Prioritize and time-phase: Build a portfolio view that balances impact, feasibility, and speed. Many programs use four categories: quick wins (low disruption), no-regrets upgrades (repeatable, short payback), scale packages (standard designs replicated across sites), and strategic redesign (capital-cycle dependent). Time-phase the portfolio so near-term milestones are covered by measures you can deliver within existing downtime and procurement lead times.
Step 8: Create a fast funding and procurement path: Efficiency loses momentum when each small project is forced through the full capital approval cycle. Establish thresholds for fast-track approvals, pre-qualify vendors, and standardize scopes for frequent measures (controls tuning, leak repair, recommissioning). For larger projects, bundle similar work to reduce transaction cost and align with maintenance shutdown windows. Decide how savings will be treated in budgets so owners are not penalized for using less energy.
Step 9: Execute with cadence and accountability: Run weekly site-level check-ins to clear blockers and manage work orders, and monthly portfolio reviews to track implementation and verified savings. Track measures through a simple lifecycle: identified, scoped, approved, implemented, verified, sustained. The lifecycle is important because it makes verification and persistence explicit deliverables, not afterthoughts.
Step 10: Verify savings and prevent drift: Verification is a control, not a celebration. Use meter data where possible and normalize for production and weather. If a measure underperforms, diagnose quickly: incorrect assumptions, poor commissioning, operator reversion, or unexpected interactions with other systems. Lock in drift prevention through alarms, automated scheduling, and periodic recommissioning for control-heavy systems. Where savings depend on behavior, document the new standard work and train operators on what to do when conditions change.
Step 11: Standardize and replicate: Scale comes from repeatable playbooks. For the highest-frequency measures, document the standard approach, typical savings range, required data, and operational guardrails. Replicate across similar sites, and use comparative performance data to improve the playbook over time.
Before launching, pressure-test readiness with a short checklist.
Mandate: Clear goals, named owners, and decision rights for operating changes.
Baseline: Bills and basic metering, plus a defined baseline period.
Drivers: Production or occupancy and weather inputs for normalization where relevant.
Pipeline: A first-wave measures list and a plan to refresh the pipeline quarterly.
Funding: Fast-track approval thresholds and pre-qualified vendors for common work.
Verification: Agreed method, cadence, and sign-off authority for savings.
Executed this way, efficiency becomes a platform. It reduces emissions now, it reduces the size and cost of later investments, and it builds the governance and data discipline you need for larger decarbonization moves.
7.5 Embedding Continuous Improvement in Operations
Efficiency savings decay unless they are protected by routines. Setpoints drift, overrides return, and maintenance slips. The antidote is to treat energy waste like a quality defect: visible, owned, and corrected quickly.
Persistence: The degree to which energy savings are maintained over time, protected from operational drift and changing conditions.
Embed energy review into existing operating rhythms. For high-energy sites, review energy intensity and top drivers weekly alongside throughput, yield, and downtime. For buildings portfolios, review night load and seasonal performance monthly. Use simple red flags to trigger investigation: rising base load, recurring peaks, or abnormal consumption not explained by weather or throughput. Institutionalize periodic recommissioning for control-heavy systems, and standardize high-efficiency specifications so replacements do not quietly degrade performance.
Build capability so energy management does not depend on a few specialists. Train operators on the “why” behind key setpoints and sequences, and document standard responses when conditions change. Align incentives so sites are not punished for saving energy (for example, by removing the ability to spend savings on other priorities). Finally, keep the baseline and verification rules stable. When people trust the numbers, efficiency becomes a management routine rather than a debate.
