Carbon Planning and MACC Curves: How Australian CFOs...

The board meets in six weeks. The chair has asked the CFO for a single slide: which emission-reduction projects should the group fund next year, and what does each dollar of capex buy in tonnes of CO2e avoided? The CFO turns to the sustainability lead. The sustainability lead has a spreadsheet from a consultant. It was built eighteen months ago. Nobody trusts it anymore.

That is the workflow we are trying to end.

Traditional capex prioritisation for emission reduction is a bespoke Excel model. A consultant builds it. It gets updated once a year if you are lucky. By month six, the ledger has moved, the emission factors have been revised, and half the assumed unit costs are stale. The CFO defaults to gut feel. The board approves the loudest project, not the most efficient one.

A Marginal Abatement Cost Curve fixes that. It ranks every candidate action by cost per tonne of CO2e avoided, over the payback horizon you choose, against the live emission ledger. Not a snapshot. Not last year's data. The actual numbers sitting behind your NGER submission today.

What a MACC curve actually is

A MACC is a bar chart. Each bar is one emission-reduction action. The width of the bar is the tonnes of CO2e that action removes per year. The height of the bar is the abatement cost, expressed in AUD per tonne of CO2e avoided over the action's economic life.

Bars are sorted left to right, cheapest first.

Negative bars sit below the x-axis. Those are actions that save more money than they cost. LED retrofits usually land there. Compressor upgrades often do too. Positive bars sit above the axis. Those are actions with a net cost. Rooftop solar without a rebate sits there for many portfolios. Refrigerant retrofits sit higher. Green hydrogen sits way, way up on the right.

The CFO reads the curve like a supply curve. Draw a horizontal line at whatever internal carbon price the organisation has set. Every bar under that line is a project the numbers already justify. Every bar above the line needs a strategic reason to fund, or a change in the input assumptions.

The maths is simple. Total capex plus opex delta over the action life, divided by total tCO2e avoided over the same life. That is your $/tCO2e. Our Carbon Planning module runs this per action and returns the sorted list via one query. On the plan-detail screen, a MACC tab renders the chart. The reporter clicks any bar to see the action detail behind it.

Where the emission-reduction actions come from

Two sources. The first is an action library. This is a repository of pre-configured action templates, tagged by category and by applicable scope. Rooftop solar sits in the renewable-energy category and applies to Scope 2. LED retrofit sits in energy efficiency and applies to Scope 2. Fleet electrification sits in transportation and applies to Scope 1. Refrigerant retrofit sits in process optimisation and covers a specific Scope 1 blind spot. Every library entry carries a typical reduction range, a typical cost range, and a typical payback range.

Reporters use library entries as starting points. Numbers get overridden with the reporter's actual quotes, real installation cost, real energy price contract.

The second source is custom actions added by the reporter or their consulting partner. A boiler electrification quote from a specific mechanical services firm. A PPA proposal from a specific retailer. A material substitution wired to the reporter's material library where the platform computes the delta emission factor per unit and multiplies by expected annual usage.

The substitution engine is worth calling out. If the action is "switch from Portland cement to a supplementary cementitious material blend", the platform pulls the current material's emission factor and the alternative's emission factor from the versioned material library, applies the reporter's actual quarterly usage, and returns the annual tCO2e reduction plus the annual cost delta. The abatement cost falls out at the bottom.

How each action carries its numbers

Every action in the ledger holds a defined set of fields. Not vibes. Actual columns in an actual table.

Estimated reduction in tCO2e per year
Estimated capex in AUD
Estimated opex savings or delta in AUD per year
Payback period in years
Return on investment as a percentage
Scope the action affects (Scope 1, 2, or a specific Scope 3 subcategory)
Start date and completion date for the implementation window
Progress percentage and actual reduction once the action is live

The abatement cost is computed. Not typed. If a reporter changes the capex figure, the abatement cost updates. If the emission factor version underlying a material substitution changes, the abatement cost recomputes. This is how the MACC stays current without a consultant rebuilding a spreadsheet.

Each action also carries an internal carbon price at the plan level. That is the reporter's chosen hurdle rate for green capex. Actions with an abatement cost below the internal carbon price should get funded on numbers alone. Actions above it need a strategic justification, which sits in the notes field and shows up in the approval workflow.

Integration with emissions forecasting

A MACC on its own tells you the cost per tonne. It does not tell you whether the actions you choose actually close the gap to your target. That is a forecasting question.

The Carbon Planning module answers it by linking every action to a plan, and every plan to one or more targets. When a reporter models a set of actions, the platform sums the projected annual reductions across the action window, subtracts them from the business-as-usual emission trajectory, and plots the resulting curve against the target line.

If the projected trajectory closes the gap, the plan is marked on-track. If it does not, the platform flags a target gap in tonnes and asks the reporter to add more actions or accept a shortfall. The plan-detail screen shows the gap explicitly. No sugarcoating.

The BAU trajectory itself is built from the emissions forecasting engine. That takes actuals from the current ledger, extrapolates them against activity drivers, and produces a monthly projection through the target year. If our emissions forecasting says next year lands at 47,200 tCO2e without intervention, and the target is 42,000, the plan needs at least 5,200 tonnes of committed action, minus a prudence buffer.

Integration with the Safeguard Mechanism trajectory check

For NGER Safeguard facilities, this stops being a strategic exercise and becomes a compliance requirement.

The Safeguard Mechanism baseline declines by 4.9% per year for most facilities. Miss the baseline and you either surrender ACCUs or pay a Safeguard Mechanism Credit shortfall. At current ACCU prices in the mid-$30s to low-$40s AUD per tonne, a 10 kt overshoot costs somewhere between $350K and $420K a year, and that is before the strategic risk of an SMC issuance decision going against you. Get this wrong for three years and the number gets attention on the P&L.

The Carbon Planning module runs a Safeguard trajectory check as a first-class filter. Point the plan at a Safeguard-covered facility, set the current-year declining baseline, and the check returns two lists. Actions that close the projected baseline gap for the current year. Actions that do not.

The CFO does not need to sort through fifty candidate actions and work out which ones move the Safeguard needle before 30 June. The trajectory check does that. The remaining question becomes: of the actions that close the gap, which have the lowest abatement cost? The MACC answers that. So the workflow is Safeguard-filter first, then MACC-sort second. The Safeguard trajectory check walk-through covers this in detail.

Scenario builder: 1.5, 2, and 3 degree pathways with different action mixes

A single plan can hold multiple scenarios. Business-as-usual is one. Planned actions is another. Ambitious is a third. Target-driven is a fourth.

Each scenario is a specific set of actions with specific implementation years and specific probability weights. A reporter models a 1.5°C-aligned scenario by including the full action list at 100% probability. A 2°C scenario might drop the highest-cost actions and keep the ones with abatement cost under $80/tCO2e. A 3°C scenario might drop everything except the actions with a negative abatement cost, the ones that pay back on operating savings alone.

The scenario builder is not a black box. Each scenario carries a projected reduction, a total cost, a target gap, and a confidence level. The confidence level is a number between 0 and 1 that the reporter sets to reflect how firm the underlying quotes and assumptions are. A rooftop solar action with three signed installer quotes carries confidence closer to 1. A boiler electrification action based on a one-page vendor estimate carries confidence closer to 0.6. The scenario's projected reduction is weighted accordingly.

We think that confidence field matters more than the pretty chart. It forces the reporter to be honest about which numbers are firm and which are directional. When the AASB S2 auditor asks how the transition plan disclosure was built, that number is the answer.

The board pack view

The Carbon Planning module renders a MACC chart designed for the board pack, not the analyst screen.

Actions are colour-coded by scope. Scope 1 in one colour, Scope 2 in another, Scope 3 in a third. Bars are grouped by payback bucket. Actions with payback under three years get a solid fill. Actions with payback three to seven years get a hatched fill. Actions with payback beyond seven years get a lighter fill. The board can read the strategic capital-intensity of the plan at a glance.

Below the chart, the board pack view lists the top ten actions with columns for capex, annual tCO2e avoided, payback in years, and abatement cost in $/tCO2e. Cumulative capex and cumulative tonnes avoided are calculated across the top ten so the CFO can say "our proposed FY27 green capex is $8.4M and buys us 14,300 tonnes a year".

The chart and the list export to PDF and Excel via the standard report engine. No copy-paste. No formatting drift.

How the CFO uses this in a capex prioritisation meeting

The workflow reduces to this.

Open the plan for the coming financial year.
Filter the action list to those that pass the Safeguard trajectory check for facilities in scope.
Sort what remains by abatement cost, ascending.
Draw a horizontal cutoff line at the internal carbon price. Fund everything below it as a default.
For actions above the cutoff, ask whether there is a strategic reason (customer contract, regulatory anticipation, reputational hedge) that justifies funding despite the number.
Lock the resulting action list as the approved plan. Approval kicks off a Segregation of Duties workflow: the plan creator cannot approve their own plan. This matters when the auditor asks about internal controls over the transition plan disclosure.

We built this to end the annual scramble where the sustainability team, the CFO office, and the operations lead argue over which projects are "priorities" without agreed numbers. The MACC produces a defensible ranking. The internal carbon pricing discussion decides where the funding cutoff falls. Those are separate decisions and should be argued separately.

The output feeds directly into the AASB S2 transition plan disclosure. Paragraphs 33 to 36 of AASB S2 require companies to disclose targets, planned actions, milestones, and progress. The Carbon Planning module generates a disclosure block covering objectives, milestones, carbon-credit reliance, and the gross-versus-net split of the reduction pathway. That block drops into the disclosure report auto-generator without retyping. The CFO playbook covers how this reads to the audit committee.

What the module does not do

Two honest limitations.

It does not automate action selection. The MACC ranks. It does not decide. A reporter can filter, sort, and rank, but a human approves the plan. The Segregation of Duties enforcement makes sure of that. Automation of the funding decision is not appropriate; the accountability sits with the CFO and the board.

It does not integrate with financial planning software. There is no direct push to Oracle EPM, Anaplan, Workday Adaptive, or Board. The abatement cost figures and cumulative capex numbers export to Excel and PDF. Someone still copies them into the FP&A model. We do publish outbound webhooks and API keys so a customer's data team can wire the flow themselves, and external AI assistants like ChatGPT or Claude can query the live plan ledger via the platform's assistant integration. But there is no native financial-planning connector today. Building one is a roadmap item, not a shipped one.

FAQ

How much emission factor precision goes into the MACC calculation?

The reduction side of every action uses versioned emission factors from the material library, snapped to the effective date of the action's implementation window. If the reporter is modelling a fleet electrification action starting in January 2028, the calculation uses the NGA Factors edition current for that period, not today's edition. If the factor is later revised, the platform recomputes the delta and updates the abatement cost. The cost side is whatever the reporter enters. It is not modelled.

Can we run one plan across multiple business units and multiple facilities?

Yes. A plan can cover a single business unit or roll up across the organisation. Actions can be tagged to specific business units. The Safeguard trajectory check runs per facility. The MACC can be filtered by business unit at render time. For joint ventures, plan rollup runs across JV partners using the chosen consolidation approach (operational control, financial control, or equity share).

What happens when an action underperforms the projected reduction after it goes live?

A daily background scan compares actual emission reductions to the projected trajectory. If the divergence exceeds the alert threshold, the platform opens a variance alert against the action. The alert flows to the action owner and to the plan approver. Alerts have a lifecycle: open, acknowledged, resolved, or dismissed with a note. That log becomes part of the audit trail on the plan.

Does the MACC support intensity targets, not just absolute reductions?

Yes. Actions can be linked to an intensity target (tCO2e per unit of revenue, per FTE, per floor area, per unit of production, or a custom denominator). The abatement cost calculation stays in $/tCO2e absolute. The progress calculation runs on intensity. The plan-detail screen shows both.

How does this compare to a consultant-built Excel MACC?

A consultant model captures a moment. This one lives with the ledger. When new invoices come in and get processed by the document engine, target progress recalculates against the ledger. When the reporter changes an assumption, the curve is recomputed from the current inputs. That is the difference between a spreadsheet and a system. The spreadsheet loses trust at month six. The system does not.

Consulting partners are the natural buyers of that system, not the competition. A practice leading an ASX200 transition plan engagement uses the platform as the workshop equipment; the consultant remains the craftsperson who chooses the actions, sets the internal carbon price, and defends the plan to the board.

Carbonly pricing is per-project with a $100/month platform fee. To talk about running your own Carbon Planning workflow against your live emission ledger, email hello@carbonly.ai.

Carbon Planning and MACC Curves: How Australian CFOs Prioritise Emission-Reduction Capex