4. Carbon Accounting, Metrics and Sustainability Analytics

Introduction and Context

Logistics creates value by moving, storing and coordinating goods, but these activities also generate greenhouse-gas (GHG) emissions. Fuels burned by trucks, ships, aircraft and material-handling equipment, electricity used in warehouses, refrigerants released from cooling systems, and emissions embedded in purchased transport and packaging all contribute to an organisation’s climate impact. Carbon accounting converts these physical activities into quantified emissions, allowing managers to establish a baseline, identify priorities, evaluate alternatives and report progress.

Carbon accounting is not simply the preparation of an annual emissions figure. It is a management system linking organisational boundaries, activity data, emission factors, data quality, operational decisions and strategic targets. In logistics, this is particularly demanding because supply chains cross organisational boundaries and involve multiple transport modes, subcontractors, suppliers and customers. A credible accounting system must therefore be technically rigorous while remaining useful for operational decision-making.

This chapter develops an applied approach to Scope 1, Scope 2 and Scope 3 accounting, with emphasis on logistics activities. It also examines performance indicators, baselines, targets, dashboards, benchmarking, marginal abatement cost curves, scenario analysis and uncertainty. These tools enable organisations to move from reporting emissions to managing them.

Learning Objectives

After studying this chapter, the learner should be able to:

1. Explain the purpose, boundaries and principles of organisational carbon accounting.

2. Distinguish Scope 1, Scope 2 and Scope 3 emissions in logistics systems.

3. Apply activity-based and spend-based methods using appropriate emission factors.

4. Establish a defensible emissions baseline and select meaningful logistics KPIs.

5. Design dashboards and benchmarking systems for sustainability performance.

6. Interpret marginal abatement cost curves and compare decarbonisation options.

7. Use scenario analysis and uncertainty assessment in logistics decisions.

8. Evaluate governance, data, ethical and managerial challenges in carbon reporting.

The Architecture of Carbon Accounting

The most widely used organisational framework is the GHG Protocol Corporate Standard. It classifies direct and indirect emissions into three scopes. Scope 1 covers direct emissions from sources owned or controlled by the reporting organisation. For a logistics company, examples include diesel burned by company-owned trucks, natural gas used in a warehouse boiler, fuel used by owned vessels and refrigerant leakage from company-controlled refrigeration equipment.

Scope 2 covers indirect emissions from purchased or acquired energy, principally electricity, steam, heating and cooling. A distribution centre that buys electricity does not directly burn fuel to generate that electricity, but its operations cause emissions at the power station. Scope 2 reporting commonly distinguishes location-based emissions, calculated using average grid emission factors, from market-based emissions, calculated using contractual instruments such as supplier-specific factors or renewable electricity certificates where applicable.

Scope 3 includes other indirect emissions occurring in the organisation’s value chain. Relevant logistics categories may include purchased transportation and distribution, upstream fuel- and energy-related activities, capital goods such as vehicles and warehouses, waste generated in operations, employee commuting, downstream transportation, use of sold products and end-of-life treatment. The precise categories depend on the organisation’s activities and reporting boundary.

The scopes should not be interpreted as three independent environmental impacts. They describe different relationships between the reporting organisation and emission sources. A retailer using a contracted carrier may report the carrier’s transport emissions in Scope 3, while the carrier reports the same vehicle fuel emissions as Scope 1. This is not necessarily double counting in a global inventory; it reflects different corporate perspectives. However, organisations must avoid adding overlapping categories within their own inventory.

Organisational and Operational Boundaries

Before collecting data, the organisation must decide which entities and activities are included. Under an equity-share approach, emissions are allocated according to ownership interest. Under control approaches, the organisation accounts for emissions from operations or assets it controls. A parent company with leased warehouses, joint ventures and outsourced transport must document how these arrangements are treated.

Operational boundaries then classify emission sources within the selected entities. A useful inventory generally includes stationary combustion, mobile combustion, purchased electricity, refrigerants, purchased logistics services, business travel, waste and relevant purchased materials. Boundary decisions should be consistent across years; otherwise apparent performance changes may reflect accounting changes rather than real reductions.

The Basic Calculation Logic

At its simplest, emissions are calculated as:

Emissions = Activity Data × Emission Factor

Activity data may be litres of diesel, kilowatt-hours of electricity, vehicle-kilometres, tonne-kilometres, pallet-days, kilograms of refrigerant or monetary expenditure. An emission factor expresses emissions per unit of activity, such as kilograms of carbon dioxide equivalent (kg CO2e) per litre of fuel or per tonne-kilometre.

Carbon dioxide equivalent allows different GHGs to be expressed in a common unit using global warming potential values. The selected assessment basis and factor version must be recorded because methodological updates can change reported results. Combustion emissions may be reported as tank-to-wheel emissions, while logistics comparisons may require well-to-wheel or life-cycle boundaries. These terms must not be mixed without explanation.

Measuring Logistics Emissions by Scope

Scope 1: Controlled Fleet and Facilities

For owned or controlled vehicles, fuel records are generally the preferred activity data. If a fleet consumes 100,000 litres of diesel and the applicable factor is 2.68 kg CO2e per litre, reported emissions are 268,000 kg CO2e, or 268 tonnes CO2e. Fuel-card records, telematics and depot-level invoices can improve completeness and permit analysis by vehicle, route or business unit.

Stationary combustion should be calculated from natural gas, heating oil or other fuel quantities. Refrigerant emissions require special care. Leakage can be estimated from purchase and recharge records, mass-balance methods or equipment-specific leakage assumptions. Because some refrigerants have high global warming potentials, apparently small quantities can be material.

Scope 2: Purchased Energy

Electricity emissions are calculated from metered consumption and an appropriate factor. Warehouses should ideally report electricity by site and major use, such as refrigeration, lighting, charging and automation. Location-based reporting supports assessment of physical grid dependence, whereas market-based reporting reflects contractual electricity procurement. Both methods should be transparently presented where required by the reporting framework.

Energy intensity indicators are often more useful than absolute emissions alone. Examples include kWh per order, kg CO2e per pallet handled and kg CO2e per square metre of warehouse space. Nevertheless, intensity improvement can coexist with rising total emissions when throughput grows; both absolute and intensity measures are therefore necessary.

Scope 3: The Difficult Boundary

Scope 3 often represents the largest and most uncertain part of a logistics organisation’s inventory. The preferred method for purchased transport is activity-based calculation: distance, payload, mode, fuel or energy consumption and load factor are combined to estimate emissions. A simplified tonne-kilometre approach is useful when detailed fuel data are unavailable, but it can conceal differences in vehicle efficiency, empty running, routing and temperature control.

Supplier-specific data can improve accuracy, provided the supplier explains its boundaries, factor sources and allocation method. When this is not possible, organisations may use average-data or spend-based methods. Spend-based methods multiply expenditure by an industry-average emissions factor. They are useful for screening but are sensitive to prices, inflation, currency and product mix. They should not be mistaken for precise operational measurement.

Allocation is a major issue in shared logistics. Emissions may be allocated by mass, volume, occupied capacity, distance, revenue or another causal driver. The selected rule should reflect the service being measured and be applied consistently. For example, allocating refrigerated transport only by shipment mass may understate the influence of temperature requirements and dwell time.

From Inventory to Management Metrics

A carbon inventory answers the question, “How much was emitted?” Sustainability analytics asks additional questions: Where did emissions arise? What caused the change? Which intervention is most effective? What will happen under alternative operating conditions? Metrics must therefore connect environmental performance with logistics outcomes.

Useful indicators include:

• total tonnes CO2e by scope, site, customer, mode and activity;
• kg CO2e per tonne-kilometre, shipment, order, pallet or revenue unit;
• fuel consumption per vehicle-kilometre and empty-running percentage;
• warehouse kg CO2e per order or per pallet handled;
• percentage of transport activity using lower-emission modes;
• renewable electricity share;
• percentage of emissions based on primary supplier data;
• emissions reduction against baseline and target trajectory.

No single KPI is sufficient. A carrier may reduce emissions per tonne-kilometre by increasing vehicle utilisation while total emissions rise because it handles more freight. Conversely, absolute emissions may fall because business volume has contracted. A balanced scorecard should therefore include absolute, intensity, activity, quality and outcome measures.

Baselines and Targets

A baseline is a defined reference year or period against which performance is evaluated. It should represent normal operations, use a documented boundary and contain sufficiently reliable data. Organisations should recalculate a baseline after structural changes such as acquisitions, divestments, outsourcing or major methodological revisions when these materially distort comparisons.

Targets may be absolute, intensity-based or hybrid. An absolute target controls total atmospheric emissions; an intensity target relates emissions to output and may be more practical during growth. Targets should specify scope, boundary, year, metric, deadline and treatment of offsets. A credible target is supported by an implementation plan rather than expressed only as an aspiration.

Analytical Tools for Decision-Making

Dashboards and Data Architecture

A sustainability dashboard should allow users to move from an aggregate result to its operational explanation. A senior dashboard may show total emissions, progress against target, emissions by scope and major hotspots. A transport manager may need route-level fuel use, load factor, idling, mode share and emissions per shipment. A data steward requires completeness, factor version, estimation percentage and unresolved anomalies.

Effective systems integrate enterprise resource planning, transport-management systems, warehouse-management systems, utility meters, telematics, fuel cards and supplier portals. Automated data collection reduces manual error, but automation does not remove the need for validation. Rules should identify missing distances, improbable fuel consumption, duplicate invoices and sudden factor changes.

Benchmarking

Benchmarking compares performance across sites, carriers, routes, modes or time periods. Internal benchmarking is often more reliable than external comparison because boundaries and data definitions are more consistent. External benchmarking can provide strategic context but requires careful examination of geography, load factor, vehicle type, service level and accounting method.

A fair benchmark compares like with like. Comparing an express airfreight service with a consolidated ocean service solely on kg CO2e per tonne may be misleading because speed, reliability and product value differ. Benchmarking should therefore include service attributes and explain the operational drivers of variation.

Marginal Abatement Cost Curves

A marginal abatement cost curve (MACC) ranks reduction measures according to their cost per tonne of avoided emissions. For each measure:

Abatement cost per tonne = Net annualised cost ÷ Annual emissions reduction

Net cost includes capital, operating, maintenance, energy and residual-value effects. A measure with negative cost saves money while reducing emissions; examples may include route optimisation or reduced idling. Measures with positive cost require investment or may be justified by compliance, resilience, customer requirements or strategic positioning.

MACCs are decision aids, not automatic investment plans. They depend on assumptions about fuel prices, utilisation, asset life, discount rate, technology performance and interactions between measures. Installing electric vehicles may change depot electricity demand; reducing shipment frequency may alter inventory and service costs. Measures should therefore be assessed both individually and as portfolios.

Scenario Analysis

Scenario analysis explores how emissions and costs change under different assumptions. A logistics organisation might compare a business-as-usual scenario with accelerated electrification, increased rail share, renewable electricity procurement or a high-growth demand scenario. Scenarios are not forecasts; they are structured representations of plausible conditions.

A useful scenario model links demand, network design, mode, distance, load factor, energy type and emission factors. Sensitivity analysis then tests influential variables. For example, the result of a fleet-electrification scenario may depend strongly on annual kilometres, battery life, electricity carbon intensity, charging availability and payload constraints.

Uncertainty and Data Quality

Uncertainty arises from incomplete activity data, estimated distances, uncertain load factors, changing emission factors, supplier claims and allocation choices. A reported figure should not imply greater precision than the data support. Organisations can use data-quality scoring based on temporal, geographical, technological and completeness dimensions. Primary, measured and supplier-specific data normally receive higher confidence than spend-based estimates, though primary data are not automatically correct.

Sensitivity analysis identifies variables that materially affect conclusions. Where appropriate, inventories can present ranges rather than a single point estimate. Improvement plans should prioritise uncertainty that could change a decision, not merely uncertainty that is easy to reduce. Data governance should record source, owner, date, unit, factor, conversion, estimation method and review status for every material dataset.

Applied Case: NorthStar Distribution Network

Background

NorthStar Distribution is an illustrative regional logistics provider operating 80 diesel trucks, three distribution centres and a contracted line-haul network. It serves retail and industrial customers and is experiencing pressure from customers to disclose transport emissions. Management wants a two-year reduction plan without reducing delivery reliability.

Situation and Analysis

The company establishes an operational-control boundary. Scope 1 includes company fleet fuel, natural-gas heating and refrigerant losses. Scope 2 includes purchased electricity at the three centres. Scope 3 includes contracted line-haul transport and selected purchased logistics services. Initial data show that fleet fuel and contracted transport are the principal sources, while warehouse electricity is smaller but more measurable.

The team finds that fuel-card data cover 96% of company vehicle fuel, but route distances are incomplete and subcontractor data are largely spend-based. It creates a baseline using fuel records for the owned fleet, utility invoices for facilities and tonne-kilometre estimates for contracted transport. It labels the results by data quality rather than presenting all figures as equally reliable.

Operational analysis identifies high empty-running on return routes, excessive idling at two depots and low utilisation on several urgent deliveries. The dashboard combines total tonnes CO2e with emissions per delivery, vehicle utilisation, empty kilometres, on-time performance and estimated-data percentage.

Strategy and Solution

NorthStar evaluates four measures: driver training and anti-idling controls, route and backhaul optimisation, renewable electricity procurement, and phased deployment of electric rigid trucks on predictable urban routes. A MACC shows that anti-idling and backhaul measures have low or negative net cost. Electric trucks require higher capital expenditure but have a substantial reduction potential where annual mileage and charging access are adequate. The company therefore sequences the measures rather than selecting only the cheapest option.

For Scope 3, NorthStar introduces a supplier data template requiring transport distance, fuel or energy use, payload, vehicle type, allocation rule and factor source. Until supplier-specific data improve, the company reports estimates separately and avoids claiming precise customer-level reductions. It establishes an absolute reduction target for controlled operations and an intensity target for contracted transport while maintaining service-level safeguards.

Outcome and Learning Points

The illustrative case demonstrates that accounting creates value when connected to operational causes. The organisation does not begin with an expensive technology purchase. It first improves the baseline, identifies controllable losses, tests alternatives and builds supplier capability. It also recognises that data quality is part of the performance system. A reduction claim supported by weak or inconsistent boundaries may damage credibility even when the operational intervention is sound.

Key lessons are that boundaries must precede calculations, intensity metrics must be accompanied by absolute totals, supplier engagement is essential for Scope 3, and decarbonisation measures should be evaluated for cost, service, feasibility and interaction effects.

Challenges, Governance and Professional Practice

Common mistakes include confusing Scope 1 with all emissions from owned products, using spend-based estimates without disclosing their limitations, changing boundaries without recalculating the baseline, comparing incompatible transport services and reporting renewable-energy purchases as if they eliminated all physical grid emissions. Another error is counting offsets as operational reductions. Offsets may have a role in residual-emission strategies, but they do not replace direct abatement and must be transparently separated from inventory performance.

Carbon data require governance comparable to financial data. Roles should be assigned for data ownership, factor approval, estimation, internal review and external assurance. Records must be auditable. Management should establish controls over units, conversion factors, version changes and supplier submissions.

Ethically, organisations should avoid selective disclosure, unsupported green claims and the transfer of emissions to suppliers merely to improve their own Scope 1 profile. Outsourcing a transport activity does not make its climate impact irrelevant. Professional reporting should communicate uncertainty, methodological choices and trade-offs, including situations where a lower-carbon option affects cost, employment, resilience or customer service.

Digital technologies can improve measurement through telematics, smart meters, Internet of Things sensors, cloud platforms and machine-learning anomaly detection. Artificial intelligence can estimate missing routes, forecast demand and optimise loads, but generated estimates require validation and explainability. Poor-quality source data can be automated at scale, and opaque algorithms may create biased allocations or untraceable claims. Technology should support, not replace, accounting judgement.

Current Developments and Best Practices

Carbon accounting is moving toward more granular, near-real-time and product-level information. Customers increasingly request shipment-level emissions data, while regulators and assurance providers expect stronger evidence, consistent boundaries and transparent methodologies. Digital product passports, automated supplier data exchange and better electricity-intensity information may improve traceability. At the same time, organisations are being challenged to distinguish real reductions from accounting changes, contractual instruments and unverified claims.

A mature logistics carbon-management system should:

• define organisational and operational boundaries before data collection;
• apply recognised standards and document all assumptions;
• prioritise measured, activity-based and supplier-specific data;
• disclose estimates, exclusions and uncertainty;
• combine absolute and intensity KPIs;
• link dashboards to operational responsibility;
• recalculate baselines after material structural changes;
• evaluate measures using cost, emissions, service and resilience criteria;
• engage suppliers rather than merely reclassifying emissions;
• maintain auditable records and independent review for material claims.

Chapter Summary

Carbon accounting translates logistics activity into comparable GHG emissions and provides the evidence base for sustainability decisions. Scope 1 covers controlled direct emissions, Scope 2 covers purchased energy and Scope 3 covers other value-chain emissions. Reliable inventories depend on clear boundaries, suitable activity data, transparent emission factors, consistent allocation and explicit uncertainty.

Effective sustainability analytics goes beyond total tonnes CO2e. Baselines, absolute and intensity KPIs, dashboards and benchmarks reveal operational drivers. Marginal abatement cost curves compare reduction options, while scenario and sensitivity analysis test strategic choices under changing demand, technology and energy conditions. Governance, ethical disclosure, supplier engagement and data quality are essential to credible reporting. The central managerial task is to connect accounting results with practical interventions that reduce emissions without ignoring service, cost, resilience and organisational context.

Key Terms

Scope 1: Direct GHG emissions from sources owned or controlled by an organisation.

Scope 2: Indirect emissions associated with purchased electricity, heating, steam or cooling.

Scope 3: Other indirect emissions occurring across an organisation’s value chain.

Carbon accounting: Systematic measurement and reporting of GHG emissions.

Activity data: Quantified operational data used in an emissions calculation.

Emission factor: Emissions released per unit of activity.

CO2e: A common unit expressing different GHGs according to their warming effect.

Organisational boundary: The entities included in an inventory.

Operational boundary: The emission sources and activities included in an inventory.

Baseline: A defined reference period used to assess change.

Intensity metric: Emissions expressed relative to an output or activity measure.

Location-based Scope 2: Electricity emissions calculated using average grid factors.

Market-based Scope 2: Electricity emissions calculated using contractual energy factors.

Tonne-kilometre: Movement of one tonne of freight over one kilometre.

Marginal abatement cost curve: A ranking of reduction measures by cost per tonne avoided.

Scenario analysis: Evaluation of outcomes under alternative assumptions.

Data quality: The reliability, completeness and suitability of emissions information.

Allocation: Assignment of shared emissions to products, customers or activities.

Carbon dashboard: A visual system for monitoring emissions and related performance indicators.

Uncertainty: The degree to which an emissions estimate may differ from the actual value.

Review Questions

Short Answer Questions

1. Distinguish Scope 1, Scope 2 and Scope 3 emissions using logistics examples.

2. What is the basic equation used in activity-based carbon accounting?

3. Why should absolute and intensity emissions metrics be used together?

4. What is the purpose of a baseline recalculation?

5. Define a marginal abatement cost curve.

Long Answer Questions

1. Explain how organisational and operational boundaries influence a logistics GHG inventory.

2. Discuss the strengths and weaknesses of activity-based, average-data and spend-based Scope 3 methods.

3. Explain how a sustainability dashboard can support different managerial levels.

4. Analyse the role of uncertainty assessment in carbon-related decision-making.

5. Discuss governance and ethical requirements for credible logistics carbon reporting.

Analytical / Application Questions

1. A carrier reports lower kg CO2e per delivery but higher total emissions after a major increase in freight volume. Analyse whether its performance has improved and identify the additional metrics required.

2. Design a two-year carbon-accounting and reduction plan for a retailer that owns warehouses but outsources all transport.

3. A company must choose between route optimisation, renewable electricity and electric vehicles. Explain how a MACC, scenario analysis and service-level constraints should be combined to make the decision.

22. Multiple Choice Questions with Answers

Question 1

A retailer contracts an external carrier to transport its goods. From the retailer’s perspective, the carrier’s transport emissions are generally classified as:

A. Scope 1

B. Scope 2

C. Scope 3

D. Outside the GHG inventory

Correct Answer: C. Scope 3

Question 2

Which calculation best represents an activity-based emissions estimate?

A. Revenue multiplied by an industry-average factor

B. Litres of fuel multiplied by an appropriate fuel emission factor

C. Number of employees multiplied by warehouse area

D. Total profit divided by total shipments

Correct Answer: B. Litres of fuel multiplied by an appropriate fuel emission factor

Question 3

Why should a logistics organisation report both absolute emissions and emissions intensity?

A. Intensity always replaces the need for total emissions

B. Absolute emissions show scale, while intensity relates emissions to operational output

C. Absolute emissions are used only for Scope 2

D. Intensity eliminates uncertainty in the inventory

Correct Answer: B. Absolute emissions show scale, while intensity relates emissions to operational output

Question 4

Which data source would normally provide the strongest basis for calculating emissions from an owned truck fleet?

A. Corporate revenue

B. Industry-average transport expenditure

C. Fuel-card records linked with fleet activity

D. The number of customers served

Correct Answer: C. Fuel-card records linked with fleet activity

Question 5

A marginal abatement cost curve primarily ranks measures according to:

A. Total company revenue generated

B. Cost per tonne of emissions avoided

C. Number of employees required

D. Annual freight volume only

Correct Answer: B. Cost per tonne of emissions avoided

Question 6

Which situation most clearly requires careful baseline treatment?

A. A routine monthly fuel purchase

B. A change in office stationery supplier

C. A major acquisition that materially changes organisational boundaries

D. A normal seasonal variation in deliveries

Correct Answer: C. A major acquisition that materially changes organisational boundaries

Question 7

Why can spend-based Scope 3 calculations be misleading?

A. They never require emission factors

B. They are unaffected by inflation or price changes

C. They may reflect expenditure and price changes rather than physical logistics activity

D. They measure only Scope 1 emissions

Correct Answer: C. They may reflect expenditure and price changes rather than physical logistics activity

Question 8

A dashboard shows falling emissions per pallet but rising total warehouse emissions. The most reasonable interpretation is that:

A. The data must automatically be incorrect

B. Efficiency improved while throughput or activity increased

C. Scope 2 has been eliminated

D. The baseline no longer matters

Correct Answer: B. Efficiency improved while throughput or activity increased

Question 9

Which practice best addresses uncertainty in supplier transport data?

A. Presenting all estimates with excessive decimal precision

B. Separating estimated data, documenting assumptions and improving supplier data collection

C. Excluding all outsourced transport

D. Replacing activity data with revenue without disclosure

Correct Answer: B. Separating estimated data, documenting assumptions and improving supplier data collection

Question 10

What is a central ethical risk when a firm outsources transport and reports only its direct fleet emissions?

A. It may overstate warehouse electricity use

B. It may conceal relevant value-chain impacts and create a misleading impression of improvement

C. It necessarily violates the definition of Scope 2

D. It makes all carbon accounting impossible

Correct Answer: B. It may conceal relevant value-chain impacts and create a misleading impression of improvement