Cost Versus Carbon in 2026: The Exchange Rate Your TMS Uses When They Disagree

A carbon exchange rate in transport planning is the monetary value a system places on one kilogram of CO2e when it must choose between a cheaper plan and a cleaner one. Every transportation management system that optimises for cost and emissions together uses one, whether or not anybody has set it, because a solver cannot rank two plans on two objectives without an implicit conversion between them. Where no number has been agreed, that rate is zero, and the platform silently resolves every conflict in favour of cost. Locus, the world’s first Decision-Intelligent, Agentic TMS, makes the rate an explicit, governed policy input rather than a property of the solver, so the trade-off is decided by the enterprise rather than by a default.

Why the Exchange Rate Matters: The Business Case

Most enterprises have never set this number. CDP’s 2025 analysis finds that only 19% of corporates are putting an internal price on environmental externalities, the majority of which is carbon. The remaining four fifths are running optimisation software that resolves cost against emissions on their behalf, using a rate nobody chose.

Among the minority who have set one, there is no consensus about what it should be. CDP reports that about half of disclosed prices fall between $9 and $124 per tonne, a fourteenfold spread across the interquartile range alone. Two comparable operators can therefore run identical fleets on identical software and produce materially different plans, without either being wrong.

The technical reason the conflict exists at all is that cost and emissions respond differently to the same lever. Fuel burn is a function of mass, and the UK Government’s greenhouse gas conversion factors methodology, drawing on the EU ARTEMIS project, shows the largest articulated vehicles emitting up to 25% more CO2 per kilometre at full load than at half load, with the same magnitude of reduction when empty and the effect linear in between. Removing a vehicle removes its fixed cost outright but only part of its emissions, so the two objectives move at different rates whenever capacity changes.

The stakes are set by where the money already sits. With last mile running 60% to 70% of total parcel delivery cost, a rate applied across daily planning touches the largest cost line in the operation, every day, without review.

What makes this hard to notice is that nothing in a standard reporting pack exposes it. An emissions inventory records what was emitted, not what could have been avoided and was not, so a plan that declined a cheap reduction and a plan that had none available produce identical entries. The decision leaves no trace in the numbers the board sees, which is why an operation can run for years on an unexamined default without anyone encountering evidence that a choice is being made at all.

How a TMS Resolves Cost Against Carbon

1. Both objectives are computed for every candidate plan

Cost aggregates driver time, fuel or energy, vehicle fixed cost and any overtime. Emissions aggregate distance against a factor that varies by vehicle class and load. Both are properties of the same plan, computed from the same route geometry.

2. The solver needs a single number to rank on

Optimisation minimises one quantity. Two objectives must be collapsed into one, either by weighting them into a combined score or by fixing one as a constraint and optimising the other.

3. Weighting converts kilograms into currency

The weight applied to emissions in that combined score is the exchange rate. Setting it to 0.10 means the system will accept up to €0.10 of extra cost to avoid one kilogram of CO2e, or €100 per tonne, and will decline any reduction priced above that.

4. An unset weight is not a neutral weight

A platform with no configured emissions weight does not balance the objectives. It applies a weight of zero, which means it will decline a plan that removes a tonne of CO2e for one euro. The default is a decision, not an abstention.

5. The rate only binds on the plans where the objectives disagree

On most days the cheapest plan is also close to the cleanest, because both reward shorter distances and fuller vehicles. The rate changes nothing on those days. It changes the plan only where the frontier between the two is steep.

6. The chosen plan carries the rate into execution

Once the plan is selected the trade-off is already made, and it is made again tomorrow. This is why the rate belongs in policy, reviewed on a governance cycle, rather than in a planner’s judgement on the morning.

What the Model Shows About the Days They Disagree

The following is an illustrative model with stated inputs rather than observed customer data. It simulates 200 operating days of depot-based regional distribution, roughly 118 stops a day across an urban core and a periphery, served by a mixed fleet of 12-tonne rigid diesel vehicles, diesel vans and range-limited electric vans. Emission factors vary linearly with load following the structure published in the UK conversion factors methodology. Routes are built with a savings heuristic under capacity and duration constraints, and for each day the full cost and carbon frontier is traced across route counts and vehicle assignments.

The opportunity is concentrated to an extreme degree. At an internal price of €0.25 per kilogram, which is €250 per tonne and therefore sits above the upper quartile of the internal carbon prices companies disclose to CDP, 77% of modelled days offered no emissions reduction worth buying. The ceiling was set deliberately high so the finding could not be dismissed as the product of a stingy price. The best 10% of days delivered 67% of the entire year’s available saving, the best 20% delivered 94%, and the bottom half of days contributed nothing at all. The trade-off is not a daily tension. It is a rare event that matters enormously when it occurs.

The concentration survives a different price. At €0.10 per kilogram, 89% of days were idle and the best fifth delivered everything. At €2.00 per kilogram, eight times the base case, a quarter of days were still idle and the best fifth still delivered 65%. The shape is a property of the operation rather than of the price chosen.

The opportunity days could not be predicted. This contradicted the expectation the model was built to test, which was that high-opportunity days would be unusual days. They were not. High-opportunity days and ordinary days were statistically indistinguishable on every demand feature examined: order volume, stop count, average distance from the depot, dispersion, and the share of stops in the periphery all correlated with available abatement at 0.06 or below in absolute terms, and the two groups differed by no more than 3% on any feature mean. The caveat is that the model varies demand within a stable operating pattern, so it does not rule out seasonal structure. What it rules out is reading tomorrow’s opportunity off tomorrow’s order book.

Setting the rate too high costs more than setting it too low. Holding the €0.25 ceiling produced a 4.5% annual emissions reduction at a blended €0.11 per kilogram, equivalent to €108 per tonne and around 0.10% of transport spend. Running the carbon-optimal plan every day instead produced a 20.5% reduction at a blended €19.01 per kilogram. That is 4.5 times the carbon for 791 times the money, because the final 78% of the theoretical reduction costs €24.39 per kilogram. That is roughly €24,400 per tonne, some two hundred times the upper quartile of disclosed internal carbon prices, and no governance process that reviewed the figure in those units would approve it.

Shadow Price and Hard Cap: Key Differences

What to Look for in Multi-Objective Optimisation

The rate is a configuration value, not a code path

The trade-off should be visible and changeable by the people accountable for it, versioned like any other policy. If adjusting it requires a vendor release, the enterprise does not control its own carbon position. The question to put to a vendor is direct: show where the number lives, who can change it, and what the system did the last time it was changed.

Emissions are evaluated before the plan is chosen

A platform that computes carbon after routing can report but cannot trade. The emissions figure must exist for every candidate plan at the point of selection, not only for the plan that was run.

Load-dependent emission factors rather than flat rates per kilometre

A flat factor per kilometre makes consolidation look better than it is, because it credits removed vehicles with emissions that mostly move to the remaining ones. The factor has to vary with payload for the trade-off arithmetic to be correct.

The decision is recorded with its inputs

When a plan costs more than the cheapest available option, the record should show the emissions avoided, the rate applied and the resulting valuation. This is what turns an environmental claim into something an auditor can follow.

Constraints and objectives are held apart

A binding external limit, such as an urban access rule, is a constraint and must make a plan infeasible rather than merely expensive. Encoding a hard limit as a heavy penalty produces plans that quietly breach it when the penalty is outweighed.

Carbon-Aware Planning in Action

A Fortune 50 parcel and logistics network. More than a million freight shipments a year across 51 sites in a 120-country network, with dispatch decided locally and no shared view of capacity. Centralising the decision raised weekly execution from 75% to 92% and surfaced more than $14M in unused capacity, including $565K at a single site. Unused capacity is also unpriced carbon, because the same underloaded vehicles carry the payload penalty on every kilometre they run.

A leading North American retailer. Ocean, rail and road ran through six legacy systems, so no plan was ever evaluated across modes. Consolidation delivered more than $1M in savings with exceptions resolved in under two hours and 95%+ route compliance. A trade-off cannot be made between options the planning system cannot see together, which is why mode coverage precedes any carbon weighting.

A global FMCG distribution network. Operations across ten Asian countries with more than 1,000 distributors and 5,000+ riders, where 12,000+ trips a month were eliminated outright against $4B+ in optimised orders. Trip elimination is the one intervention where cost and carbon never conflict, because a trip that does not run costs nothing and emits nothing. It is the part of the frontier every operation should exhaust before it needs an exchange rate at all, and an operation still finding trips to remove has not yet reached the point where the trade-off question applies.

Common Mistakes in Setting a Carbon Trade-Off

Leaving the rate unset and assuming neutrality. An unconfigured weight is zero, and zero is a strong position rather than an absent one. It commits the operation to refusing every reduction that costs anything at all.

Setting it from a target rather than a budget. A rate derived by working backwards from a percentage reduction commitment will chase the expensive end of the curve on the few days that offer anything, which is how abatement programmes end up paying orders of magnitude above any disclosed carbon price. The target is a statement about outcomes and the rate is a statement about willingness to pay, and deriving one mechanically from the other removes the budget constraint entirely.

Reviewing performance as an annual average. The model shows half the days contributing nothing and a tenth contributing most of the result. An annual figure cannot distinguish a well-run policy from an unset one, because both look similar in aggregate.

Handling the trade-off through planner discretion. The days it matters are indistinguishable in advance from the days it does not, so a decision that depends on someone noticing will be missed exactly when it counts, and will not be consistent between depots when it is not. Discretion also destroys the audit trail, because a judgement made at a workstation carries no record of the valuation that justified it.

How Locus Approaches the Cost and Carbon Trade-Off

Locus, the world’s first Decision-Intelligent, Agentic TMS, treats the exchange rate as a governed policy input rather than a solver setting. The route planning and dispatch layer evaluates cost and emissions on every candidate plan, with load-dependent factors so the carbon term reflects payload rather than distance alone, and urban access rules held as constraints that make a plan infeasible rather than expensive. Because the days on which the trade-off binds cannot be identified in advance, the rate is applied automatically to every plan rather than raised for review on the days somebody notices. Six governance mechanisms covering explainability, traceability, evaluation, autonomy levels, execution sandbox and human-in-the-loop record the emissions avoided, the rate applied and the resulting valuation for each decision, and the Control Tower carries that record through execution so the plan that ran can be compared with the plan that was priced.

The platform reasons across more than 250 real-world operating constraints over 1.5B+ deliveries for 360+ enterprise customers in 30+ countries at 99.99% uptime, with $320M+ in aggregate logistics cost savings, 800M+ miles reduced and 17M+ kg of CO2 avoided. Locus has been recognised by Gartner for seven consecutive years, including the 2026 Gartner Hype Cycle for Supply Chain Execution and Logistics Technologies and the 2026 Gartner Market Guide for Multicarrier Parcel Management Solutions, where ShipFlex is featured as a Representative Vendor. Locus holds Leader designation in the QKS SPARK Matrix for Transportation Management Systems 2025 and the #1 position for Route Planning in G2’s 2026 Best Software Awards. In October 2025, Ingka Investments, the investment arm of Ingka Group, the world’s largest IKEA retailer, acquired Locus. Locus continues to operate independently.

Two deployments show what has to be in place before a rate can be applied at all. A Fortune 50 parcel and logistics network ran more than a million freight shipments a year across 51 sites and a 4,500-strong driver pool, with each site measuring utilisation against its own plan, so no vehicle was ever weighed against an alternative outside its depot. Centralising the decision on Locus lifted weekly execution from 75% to 92% and exposed more than $14M in capacity that had existed all along, at 99.99% uptime. A global FMCG distribution network running across ten Asian countries with 1,000+ distributors and 5,000+ riders eliminated 12,000+ trips a month against $4B+ in optimised orders, at 3X ROI. Both results came from widening the set of plans the system could compare, which is the precondition for pricing a trade-off rather than inheriting one.

Cost and carbon agree on most transport plans and diverge sharply on a few, and the number a platform uses to break the tie decides what happens on the days that matter. Modelling 200 operating days put 77% of them beyond reach at a defensible internal price, concentrated 94% of the year’s available reduction in the best fifth of days, and found no way to identify those days in advance, while pushing the rate to its limit bought 4.5 times the carbon for 791 times the money. Locus makes that rate explicit, applies it to every plan automatically because the binding days cannot be forecast, and records each trade-off with the inputs that produced it. Request a Locus carbon and cost trade-off review to see what your current planning defaults are deciding on your behalf.

Frequently Asked Questions

What is a carbon exchange rate in route optimisation? It is the monetary value a planning system assigns to avoiding one kilogram of CO2e when choosing between plans. It converts emissions into the same units as cost so a solver can rank options on a single score. Where no value is configured, the system behaves as though the rate is zero.

Do cost and carbon usually conflict in transport planning? No. Both objectives reward shorter distances and fuller vehicles, so the cheapest plan is usually close to the cleanest. In the model described here, 77% of days offered no emissions reduction worth buying at a defensible internal price, which is why the conflict is easy to overlook.

What internal carbon price do companies actually use? CDP reports that only 19% of corporates put an internal price on environmental externalities at all, and that about half of disclosed prices fall between $9 and $124 per tonne. That interquartile range alone spans a factor of fourteen, so there is no industry standard to default to.

Is a hard emissions cap better than a priced trade-off? They answer different questions. A cap suits a binding external limit, because it makes a non-compliant plan infeasible rather than merely expensive. A priced trade-off suits a voluntary target with a budget, because it buys reduction while it is affordable and stops when it is not.

Why can a flat emissions factor per kilometre distort the trade-off? Because fuel burn varies with load. UK government conversion factors show the largest vehicles emitting up to 25% more CO2 per kilometre at full load than at half load, linear in between, so a flat per-kilometre factor credits a removed vehicle with emissions that largely transfer to the vehicles that remain.

How should the trade-off be governed? As a versioned policy owned by the people accountable for both cost and emissions, applied automatically to every plan and reviewed on a fixed cycle. Leaving it to planner discretion fails because the days on which it binds cannot be distinguished in advance from the days on which it does not.