Marginal Abatement Cost (MACC) Calculator — Build & Rank Your Decarbonisation Curve
Rank decarbonisation measures by cost per tonne of CO2e avoided, build the marginal abatement cost curve, and read off which measures are economically viable below a carbon price you set.
What the calculator computes. For each measure you enter, the engine works out its marginal abatement cost — the net cost, per tonne of CO₂e avoided each year, of running that measure over its life. It then sorts every measure from cheapest to most expensive and stacks them into a curve, so you can see your whole decarbonisation portfolio ordered by cost-effectiveness and read off how much abatement is available below any carbon price.
The formula. Each measure’s cost is annualised with a capital-recovery factor and netted against its annual saving: MAC = (CRF × CAPEX − annual net saving) ÷ annual abatement, where CRF spreads the up-front capital cost evenly across the measure’s lifetime at your discount rate. A measure that saves more each year than its annualised capital costs has a negative marginal abatement cost — it pays for itself while cutting carbon.
One signed saving field, not two. Each measure takes a single annual net saving figure: energy and operating savings, net of any extra running cost, as one signed number. A positive value is a net annual saving; a negative value is a net annual cost (a measure that costs more to run than the status quo, such as a green-hydrogen switch at today’s fuel prices). There is no separate operating-cost field to reconcile.
The discount rate is a choice, and it is disclosed. The engine defaults to 3.5% — the HM Treasury Green Book social discount rate — under the Societal basis. Switch to the Private basis to appraise at a company cost of capital (an 8% preset, editable). The rate materially changes every measure’s cost, so the basis is stated in the output rather than buried.
Abatement: direct or derived. Enter each measure’s annual abatement directly in tCO₂e, or derive it from an avoided activity — kilowatt-hours of grid electricity displaced, litres of fuel no longer burnt — which the engine multiplies by the matching Ember grid or DEFRA fuel factor read live from MasterBrain and cited on the result. Measure economics — capital, savings, lifetime — are always yours; the tool structures, annualises, and ranks them but supplies no measure prices.
The carbon-price line. Set a threshold price (type it, or drag the dashed line across the curve) and the tool reports the economically viable set: how many measures sit at or below that price, the tonnes they abate, that share of your total potential, the capital required, and the net annual cost of the viable subset. In euro mode the threshold can be seeded from the latest EU ETS allowance average as an editable starting point.
HM Treasury Green Book social discount rate = 3.5% (societal). Commercial WACC is typically 7–10% (private).
Every measure’s economics are your own inputs — GreenCalculus does not supply a measure-cost library. The engine ranks measures by marginal abatement cost (cost ÷ carbon abated) and draws the curve, cheapest first.
Add measures with a cost and a carbon saving to build the curve
Results appear instantly. A ranked marginal-abatement-cost curve, the no-regret (negative-cost) pool, a draggable carbon-price threshold, a ranked table, and the full audit trail appear after calculation.
This is an indicative capital-allocation and decision-support estimate, not investment, tax, or accounting advice. Every measure’s cost, saving, lifetime, and abatement is your own input — the ranking is only as accurate as the numbers you enter. A marginal abatement cost curve ranks measures independently: it does not model interactions between measures, a changing baseline, or the order in which you implement them, and it is sensitive to the discount rate and lifetimes chosen. Derived-abatement factors are read from MasterBrain (Ember grid intensities, DEFRA fuel factors) and cited per result; the carbon-price reference, where shown, is an indicative EU ETS figure in its native currency, not a live or authoritative price. Marginal cost is not the only decision criterion — co-benefits, deliverability, capital constraints, and non-market value also matter. Validate material decisions with your own appraisal and, where required, third-party verification (e.g. ISO 14064-3).
Every organisation with a net-zero target faces the same question in a different order: not whether to decarbonise, but what to do first. Some measures save money the day they switch on. Others cost more per tonne than the entire carbon budget can bear. Without a way to rank them, capital flows to whatever has the loudest internal champion.
A marginal abatement cost curve is the ranking. It puts every option on one axis — cheapest tonne to most expensive — so the sequence of a decarbonisation programme stops being a matter of opinion.
Marginal abatement cost is the net cost per tonne of CO₂e a measure avoids: (annualised capital cost − annual saving) ÷ annual abatement. Ranking every measure by this cost builds the MACC curve — negative-cost measures pay for themselves; the rest are ranked against a carbon price.
What a marginal abatement cost curve is
A marginal abatement cost curve — MACC — is a chart that ranks emissions-reduction measures by their cost per tonne of CO₂e avoided, from cheapest to most expensive, showing the cumulative abatement available as you move up the cost ladder. Each measure becomes a block: its width is the annual tonnes it abates, its height is the net cost per tonne. Read left to right, the curve answers the two questions a decarbonisation programme actually turns on — how much carbon can I cut, and what does each successive tonne cost?
The concept has been a staple of climate-economics analysis for decades, and it has two distinct lives. As a public communication device it appears as the famous stepped chart of a whole economy’s abatement options; as an internal management tool it is a bottom-up ranking of one organisation’s own projects. This calculator builds the second kind: your measures, your costs, your curve.
You supply each measure’s capital cost, annual net saving, lifetime, and abatement. The engine does the rest: it annualises the capital, nets it against the saving, divides by the abatement to get a cost per tonne, sorts every measure cheapest-first, and stacks them into the curve. It supplies no measure prices of its own — the economics are entirely yours. Its job is to structure, rank, and let you test the portfolio against a carbon price.
How marginal abatement cost is calculated
The marginal abatement cost of a single measure is one division, built on an annualised cost:
MAC (per tCO₂e) = (CRF × CAPEX − annual net saving) ÷ annual abatement (tCO₂e/yr)
The engine spreads a measure’s up-front capital cost across its lifetime using a capital-recovery factor (CRF) — the same annuity mathematics a mortgage uses — so a £180,000 investment in a 25-year asset becomes a level annual charge rather than a lump in year one. The CRF depends on the measure’s lifetime and your discount rate:
CRF = r(1 + r)n ÷ ((1 + r)n − 1)
where r is the discount rate and n the lifetime in years. At a zero discount rate the engine falls back to straight-line annualisation (capital ÷ lifetime). The annualised capital is then reduced by the measure’s annual net saving, and the result is divided by the tonnes the measure abates each year to give a cost per tonne.
The single signed net-saving field
Each measure carries one annual net saving figure — a signed number that already combines energy savings, operating-cost changes, and any additional running cost into a single value. A positive figure means the measure saves money each year; a negative figure means it costs money to run relative to the status quo. A rooftop solar array might save £26,000 a year (positive); a green-hydrogen boiler might cost £20,000 a year more than the gas boiler it replaces (negative). There is no second operating-cost field to enter or reconcile — the sign carries that information.
When a measure’s annual net saving exceeds its annualised capital cost, its marginal abatement cost comes out negative. These are the “no-regret” measures — they save money and cut carbon, so the only reason not to do them is a lack of capital or attention, not economics. On the curve they sit below the zero line, on the far left. The engine reports them as a distinct no-regret pool: how many there are, how much they abate, and how much they save each year in aggregate.
Why measures sort by cost, not size
Once every measure has a cost per tonne, the engine sorts them ascending — cheapest first — and tracks the running total of tonnes abated. That ordering is the curve: the first block is the cheapest tonne of abatement available, the last is the most expensive, and the cumulative width tells you how much total abatement the portfolio delivers. A large measure that abates a lot but costs a fortune per tonne sits far to the right; a small, cheap measure sits at the front. Size determines a block’s width, cost determines its position.
Reading a MACC — anatomy of the curve
A marginal abatement cost curve looks unlike a normal chart, and reading it fluently is most of the value. The tool draws the curve live and lets you interrogate it; the anatomy below is what each part means.
Bar width = tonnes
The horizontal width of each block is the annual abatement that measure delivers, in tCO₂e per year. Wide blocks are big-impact measures; narrow blocks are marginal ones. The total width of the whole curve is your portfolio’s total annual abatement potential.
Bar height = £/tonne
The vertical height (or depth, below the line) is the marginal abatement cost of that measure — its net cost per tonne. Blocks below the zero line are the money-saving no-regret measures; blocks above it cost money per tonne, rising as you move right.
The carbon-price line
A horizontal dashed line at a carbon price you choose. Everything below the line is cheaper than that price — economically worth doing if carbon is valued at that level. Everything above is more expensive than the carbon price. Where the line crosses the curve is your economic cut-off.
The area, not just the height
Because width is tonnes and height is cost-per-tonne, the area of each block is its total annual cost or saving. A tall thin block can cost less in total than a short wide one. Reading area, not just height, is what separates a useful MACC reading from a misleading one.
Drag the carbon-price line across the live curve and watch the viable-set readout update: the number of measures below the line, the tonnes they abate, that share of your total potential, and the capital required. This is the fastest way to answer “what could we do for under £X a tonne?” — the question most decarbonisation budgets actually pose.
The inputs that decide the answer
A MACC is only as trustworthy as the assumptions behind each measure. Four inputs per measure, plus two global settings, determine every number the curve shows.
| Input | Unit | What it drives |
|---|---|---|
| CAPEX | Currency (≥ 0) | The up-front capital cost, annualised over the lifetime by the CRF. |
| Annual net saving | Currency/yr, signed | Energy and operating savings net of extra running cost, as one figure. Positive saves; negative costs. |
| Lifetime | Years (≥ 1) | The period the capital is spread over. Longer life → lower annualised capital → lower cost per tonne. |
| Abatement | tCO₂e/yr | The tonnes avoided each year — entered directly, or derived from an avoided activity × an emission factor. |
| Discount rate (global) | % (0–30) | Sets the CRF. Higher rate → higher annualised capital → higher cost per tonne. |
| Carbon-price threshold (global) | Currency/tCO₂e | The line the curve is tested against; defines the economically viable set. |
The discount rate: societal versus private
No single input shifts a MACC more than the discount rate, and there is no universally correct value — only a correct value for a stated purpose. The engine offers two bases. The Societal basis uses 3.5%, the HM Treasury Green Book social discount rate, appropriate for public-interest and long-horizon appraisal. The Private basis appraises at a company cost of capital — an 8% preset, editable to your own weighted-average cost of capital. A higher discount rate makes capital-heavy, long-lived measures look more expensive per tonne, because more of their up-front cost is charged to the early years.
The table below shows the same two measures appraised at three rates. A capital-heavy measure like heat pumps swings far more than a measure whose cost is dominated by savings.
| Measure | MAC at 3.5% (Green Book) | MAC at 8% (typical WACC) | MAC at 10% |
|---|---|---|---|
| Rooftop solar PV (£180k, £26k/yr saving, 60 tCO₂e, 25 yr) | −£251.3/tCO₂e | −£152.3/tCO₂e | −£102.8/tCO₂e |
| Air-source heat pumps (£240k, £12k/yr saving, 130 tCO₂e, 18 yr) | £47.7/tCO₂e | £104.7/tCO₂e | £132.8/tCO₂e |
Both stay on the same side of zero across the range, but their cost more than doubles for the heat pumps between the Green Book rate and a 10% cost of capital. A curve built at one rate should never be compared to a curve built at another — always state the basis.
Direct versus derived abatement
Abatement can be entered two ways. Direct: type the annual tonnes avoided, if you already know them. Derived: enter the activity avoided — kilowatt-hours of grid electricity displaced by solar, litres of diesel no longer burnt by an electrified fleet — and the engine multiplies it by the matching grid or fuel emission factor read live from MasterBrain, citing the exact factor on the result. Derived mode is useful when you know the energy or fuel change but not the carbon directly; it keeps the abatement figure consistent with the same factor sources used across the rest of the site.
A MACC is a model, and its ranking is only as sound as the four inputs behind each measure. Optimistic savings, generous lifetimes, or a low discount rate can flip a measure from costly to no-regret. Before trusting a curve, pressure-test the inputs: are the savings net of all extra running costs? Is the lifetime realistic or the vendor’s best case? Is the discount-rate basis (societal or private) the right one for the decision? A curve with unexamined inputs is a confident-looking answer to a question you have not actually asked.
Worked example — a six-measure abatement portfolio
This is the calculator’s built-in default scenario — open the tool and you will see these exact numbers. A mid-size site appraises six decarbonisation measures in GBP, at the 3.5% Green Book societal rate, with the carbon-price line set to £80/tCO₂e.
LED lighting retrofit £45,000 / £14,000 / 12 yr / 90 tCO₂e · Building controls (BMS) tune £30,000 / £11,000 / 10 yr / 70 tCO₂e · Rooftop solar PV 200 kW £180,000 / £26,000 / 25 yr / 60 tCO₂e · Air-source heat pumps £240,000 / £12,000 / 18 yr / 130 tCO₂e · EV fleet transition £320,000 / £18,000 / 10 yr / 150 tCO₂e · Green hydrogen boiler £600,000 / −£20,000 / 15 yr / 110 tCO₂e (the hydrogen switch costs £20,000 a year more to run than the gas boiler).
The engine annualises each measure’s capital with its CRF, nets the saving, divides by abatement, and sorts cheapest-first. The ranked curve:
| # | Measure | CRF | Annualised CAPEX | Net cost/yr | MAC (£/tCO₂e) | Cumulative tCO₂e |
|---|---|---|---|---|---|---|
| 1 | Rooftop solar PV | 0.06067 | £10,921 | −£15,079 | −£251.3 | 60 |
| 2 | Building controls (BMS) | 0.12024 | £3,607 | −£7,393 | −£105.6 | 130 |
| 3 | LED lighting retrofit | 0.10348 | £4,657 | −£9,343 | −£103.8 | 220 |
| 4 | Air-source heat pumps | 0.07582 | £18,196 | £6,196 | £47.7 | 350 |
| 5 | EV fleet transition | 0.12024 | £38,477 | £20,477 | £136.5 | 500 |
| 6 | Green hydrogen boiler | 0.08683 | £52,095 | £72,095 | £655.4 | 610 |
The three cheapest measures — solar, building controls, and LED — all have negative marginal abatement costs: together they abate 220 tCO₂e a year and save £31,815 a year. They are the no-regret pool, and they sit below the zero line at the front of the curve. Heat pumps and the EV fleet cost money per tonne but stay moderate; the green-hydrogen boiler, at £655/tonne, is an order of magnitude more expensive than anything else — the far-right block that a carbon price well below £655 leaves out.
The £80/tCO₂e carbon-price line tells the strategy story:
At £80/tCO₂e, four of the six measures are economically viable (the three no-regret measures plus heat pumps at £47.7/tonne). Those four abate 350 tCO₂e/yr — 57.4% of the total potential — for £495,000 of capital, and the viable subset is net cash-positive at −£25,619/yr (it saves money overall). The EV fleet (£136.5/tonne) and hydrogen boiler (£655.4/tonne) sit above the line: worth doing only if carbon is valued higher, or if a target requires the abatement regardless of cost.
The per-measure “net cost/yr” figures are display-rounded, so summing the visible column lands within about £1 of the £66,954 portfolio total — the engine sums unrounded intermediates and rounds once at the end.
Public vs expert MACCs — two traditions
MACCs come in two flavours, and conflating them is a common source of both over-claiming and unfair criticism. Knowing which kind you are building — and which kind a curve you are shown was built as — is essential to reading it honestly.
| Public / top-down MACC | Expert / bottom-up MACC | |
|---|---|---|
| Scope | A whole economy, sector, or country. | One organisation, site, or asset portfolio. |
| Built from | Modelled abatement potentials and generalised cost assumptions. | Actual project costs, savings, and lifetimes for real measures. |
| Strength | Communicates the big picture; frames policy debate. | Directly actionable; each block is a decision someone can make. |
| Main critique | Hides interactions, assumes independence, and can imply “free” abatement that market barriers prevent. | Only as good as its inputs; a static snapshot of a moving system. |
This calculator builds the second kind — a bottom-up, expert MACC from your own measures. The most-cited critique of MACCs (that the negative-cost “free lunch” measures are never actually free, because barriers like capital constraints, split incentives, and transaction costs keep them undone) is real, and it applies to your curve too. The no-regret pool the tool identifies is where the economics are favourable; whether those measures actually get done depends on the organisational barriers a curve cannot see.
Using a MACC in decarbonisation strategy
A curve is an input to a decision, not the decision itself. Three uses turn the ranking into a strategy.
Prioritise against the carbon-price line
The simplest use is sequencing: do the no-regret measures first (they fund themselves), then work rightward up the curve to the point where cost per tonne exceeds the value you place on carbon. The carbon-price line makes that cut-off explicit — everything below it clears an economic bar; everything above needs a different justification, such as a hard reduction target or a regulatory obligation.
Pair the curve with an internal carbon price
The carbon-price line is most powerful when it is not arbitrary. Feeding the curve with your own internal carbon price ties the viability cut-off to the price you already apply in capital appraisal — so the MACC and your investment decisions use the same value of carbon. A measure above your internal price on the MACC is a measure your own appraisal would already reject; one below it is one your appraisal would favour. The two tools share a spine: the internal carbon pricing methodology sets the price, the MACC ranks what that price makes viable.
If your internal carbon price is a long-horizon, society-facing number, appraise the MACC on the Societal (3.5%) basis so the two are consistent. If it is a hurdle-rate-style price tied to your cost of capital, use the Private basis. Mixing a societal carbon price with a private discount rate quietly biases which measures clear the line.
Know where the curve misleads
A MACC is a static snapshot of independent measures, and reality is neither static nor independent. Costs fall over time (today’s expensive measure may be cheap in five years), measures interact (insulating a building changes the heat pump you need), and the ordering ignores that some measures unlock others. Treat the curve as a structured starting point for a sequencing conversation, not a fixed programme — and rebuild it as costs and your portfolio change.
Common mistakes and analytical pitfalls
The failure modes below are the ones that most often turn a MACC from a decision aid into a misleading artefact. Most are input or interpretation errors, not arithmetic ones.
01 — Double-counting savings and running costs
The annual net saving is a single signed field: energy and operating savings net of any extra running cost. Entering the gross energy saving and forgetting the higher maintenance cost of the new kit inflates the saving and pulls the measure artificially left on the curve. Net everything into the one figure.
02 — Comparing curves built at different discount rates
A measure’s cost per tonne can double between a 3.5% and a 10% discount rate. A curve appraised on the Societal basis is not comparable to one on the Private basis. Always state the rate and basis, and never overlay two curves built at different rates.
03 — Treating the no-regret pool as automatically done
Negative-cost measures are favourable on economics alone, but capital constraints, split incentives, and organisational inertia keep many of them undone in practice. The tool identifies where the economics are good; it cannot tell you whether the barriers to acting on them have been cleared.
04 — Reading height instead of area
A tall, narrow block (high cost per tonne, few tonnes) can cost less in total than a short, wide one. Because total cost is width × height, judging a measure’s importance by its height alone misranks its budget impact. Read the area.
05 — Ignoring measure interactions
The curve assumes measures are independent, but they rarely are — building fabric changes the heat-pump size, on-site solar changes the value of electrification. A curve is a first-pass ranking; check the biggest interactions before committing capital in the curve’s order.
06 — Optimistic lifetimes
Lifetime spreads the capital: a generous life lowers the annualised cost and flatters the measure. Use realistic asset lives, not the vendor’s best case — the tool warns above 50 years for a reason.
07 — Forgetting the curve is a snapshot
Technology costs fall and carbon prices rise. A measure above the line today may be viable in three years. Rebuild the curve periodically rather than treating one appraisal as a fixed multi-year programme.
08 — Confusing abatement cost with a market carbon price
The MAC is your cost to avoid a tonne; a carbon price is what a market or regulator charges to emit one. The carbon-price line compares the two, but they are different quantities — a £655/tonne abatement cost is not “a £655 carbon price”, it is a measure a £655 carbon price would be needed to justify.
MACC in target-setting and disclosure
A marginal abatement cost analysis is not itself a required disclosure datapoint, but it is a common — and increasingly expected — piece of evidence behind the transition plans several frameworks now ask for. The crosswalk shows where the analysis feeds in.
| Framework | Where a MACC feeds in |
|---|---|
| SBTi transition planning | A validated science-based target sets the required abatement; the MACC shows the cost-ordered set of measures that can deliver it, and where the target forces action above the economic cut-off. |
| CSRD ESRS E1 transition plan | ESRS E1 requires disclosure of decarbonisation levers and the capital allocated to them. A MACC is a natural way to evidence which levers were chosen and why — the cost-effectiveness ranking behind the plan. |
| TCFD / IFRS S2 | Transition-risk and opportunity assessment draws on the cost of abatement under different carbon-price scenarios — exactly the sensitivity the carbon-price line tests. |
| CDP Climate Change | Emissions-reduction initiatives and their payback are reported in the questionnaire; the MACC’s per-measure cost and abatement are the underlying figures. |
Across all four the value is the same: a MACC turns “we have a plan” into “here is the cost-ordered evidence for the plan”, which is precisely the shift from ambition to credibility that these frameworks are trying to force.
Data, assumptions, and transparency
This calculator fabricates no measure prices. You supply every measure’s cost, saving, lifetime, and abatement; the tool structures, annualises, and ranks them. Two things — and only two — are read from MasterBrain, both cited on the result.
Derived-abatement factors
When you derive a measure’s abatement from an avoided activity, the engine reads the matching emission factor live from MasterBrain — Ember grid intensities for displaced electricity, DEFRA fuel factors for avoided combustion — and cites the exact factor key on the result. This keeps a derived abatement figure consistent with the factor sources used across the rest of the site.
An indicative carbon-price seed
In euro mode, the carbon-price threshold can be seeded from the latest EU ETS allowance average (€64.76/tCO₂e for 2024) as an editable starting point — the closest thing to a real, traded compliance-carbon price. It is a convenience default you can overwrite, not a recommended threshold; in sterling or dollar mode it appears as an indicative context note only, with no prefill and no currency conversion.
Everything else is yours or is an engine constant. The 3.5% default discount rate is the HM Treasury Green Book social rate — a stated convention, not a MasterBrain value — and the 8% private preset is a typical cost-of-capital placeholder you edit to your own figure. There is no MasterBrain price for any measure, no house cost curve, and no fabricated abatement cost: if you enter nothing, the tool computes nothing. Currency is a label only — all arithmetic is single-currency with no FX conversion, so enter every figure in one currency.
If a derived measure’s emission factor cannot be resolved from MasterBrain, that measure is excluded from the curve rather than defaulted to a made-up factor — the measures you entered directly still compute. A measure whose abatement is zero or negative is excluded before any division, so the curve never shows a nonsensical cost. The tool never fabricates a value to fill a gap, and never fails on missing data.
The full methodological detail — the capital-recovery derivation, the discount-rate conventions, the derived-abatement factor handling, and the treatment of interacting measures — is on the paired marginal abatement cost methodology page.
Frequently asked questions
Marginal abatement cost is the net cost of avoiding one tonne of CO₂e emissions through a specific measure, over that measure’s life. It is calculated as the annualised capital cost minus the annual net saving, divided by the annual tonnes abated. A measure whose annual savings exceed its annualised capital cost has a negative marginal abatement cost — it saves money while cutting carbon. Ranking every measure by this cost, cheapest first, produces the marginal abatement cost curve.
For each measure, work out its cost per tonne: annualise the capital cost with a capital-recovery factor at your discount rate, subtract the annual net saving, and divide by the annual abatement. Then sort every measure from lowest cost per tonne to highest and stack them, tracking the cumulative tonnes abated. Each measure becomes a block whose width is its annual abatement and whose height is its cost per tonne — the stacked blocks are the curve. This calculator does all of that from the measures you enter.
It means the measure saves money over its life while also cutting emissions — its annual net saving is larger than its annualised capital cost. These are the “no-regret” measures: economically worth doing on their own merits, before carbon is even priced. On the curve they sit below the zero line at the far left. The calculator groups them into a no-regret pool and reports how many there are, how much they abate, and how much they save each year in total. The catch is that “economically favourable” does not mean “already done” — capital limits and other barriers often keep no-regret measures undone.
The discount rate sets the capital-recovery factor, which spreads a measure’s up-front capital cost across its lifetime. A higher rate front-loads more of that cost into the early years, raising the annualised capital and so the cost per tonne — especially for capital-heavy, long-lived measures like heat pumps. For example, rooftop solar in the worked example moves from −£251/tonne at 3.5% to −£103/tonne at 10%. The engine defaults to 3.5% (the HM Treasury Green Book social rate) on the Societal basis, with an 8% cost-of-capital preset on the Private basis. Always state which basis a curve uses.
A capital-recovery factor (CRF) converts a one-off capital cost into a level annual charge over an asset’s life, using the same annuity mathematics as a mortgage repayment. It is CRF = r(1+r)ⁿ ÷ ((1+r)ⁿ − 1), where r is the discount rate and n the lifetime. The MACC uses it so that a large up-front investment is compared fairly against the annual savings and abatement it delivers, rather than charging the whole capital cost to year one. At a zero discount rate the engine falls back to simple straight-line annualisation (capital ÷ lifetime).
No — each measure takes a single signed “annual net saving” figure. Combine the energy saving, any operating-cost change, and any extra running cost into one number: positive if the measure saves money each year, negative if it costs money to run relative to the status quo. A green-hydrogen boiler that costs £20,000 a year more than the gas boiler it replaces is entered as −£20,000. There is no separate operating-cost field to reconcile against the saving.
You set a carbon price — type it, or drag the dashed line across the curve — and the tool reports the economically viable set: the number of measures with a cost per tonne at or below that price, the tonnes they abate, that share of your total potential, the capital required, and the net annual cost of the viable subset. Everything below the line is cheaper than valuing carbon at that price; everything above needs a different justification, such as a hard target. In the worked example, an £80/tonne line makes four of six measures viable, covering 57.4% of the total abatement potential for £495,000 of capital.
Either from you or from a factor. In Direct mode you type the annual tonnes avoided. In Derived mode you enter the activity avoided — kilowatt-hours of grid electricity displaced, litres of fuel not burnt — and the engine multiplies it by the matching Ember grid or DEFRA fuel emission factor read live from MasterBrain, citing the exact factor on the result. The measure’s economics (capital, saving, lifetime) are always yours; only the derived-abatement factor is read from the data layer, and only when you choose Derived mode.
The main critiques are that public, top-down MACCs treat measures as independent when they interact, imply that negative-cost abatement is “free” when barriers keep it undone, and present a static snapshot of a system that changes as costs fall and prices rise. Those critiques are real. A bottom-up MACC like this one — built from your own project costs — avoids some of the generalisation problems, but still shares the static-snapshot and independence limitations. Use a MACC as a structured starting point for sequencing, not as a fixed multi-year programme, and rebuild it as your costs and portfolio change.
No. Every measure’s capital cost, annual net saving, lifetime, and abatement are supplied by you — the tool holds no price for any measure and fabricates none. It reads only two things from MasterBrain, both cited: the emission factors used to derive abatement from an avoided activity, and (in euro mode) an editable EU ETS carbon-price average to seed the threshold line. The 3.5% default discount rate is the HM Treasury Green Book social rate, a stated convention rather than a MasterBrain value. If you enter no measures, the tool computes nothing.
No. Marginal abatement cost is your cost to avoid a tonne of emissions through a specific measure. A carbon price is what a market or regulator charges to emit a tonne. The MACC’s carbon-price line compares the two — it shows which of your measures cost less per tonne than a given carbon price — but they are distinct quantities. A measure with a £655/tonne abatement cost is not “a £655 carbon price”; it is a measure that only becomes economically worthwhile if carbon is valued at or above £655 a tonne.
It depends on the purpose. The Societal basis (3.5%, the HM Treasury Green Book social rate) suits public-interest, long-horizon appraisal and aligns with a society-facing carbon price. The Private basis (an 8% cost-of-capital preset, editable) suits a company appraising against its own hurdle rate. Neither is universally “correct” — but the choice should match the rest of the analysis. If you pair the MACC with an internal carbon price, use a discount basis consistent with how that price was set, or you quietly bias which measures clear the line.
Methodology notes and limitations
The formula. Each measure’s marginal abatement cost is (CRF × CAPEX − annual net saving) ÷ annual abatement, where CRF = r(1+r)ⁿ ÷ ((1+r)ⁿ − 1) annualises the capital over the lifetime at the discount rate. At a zero rate the engine uses straight-line annualisation (capital ÷ lifetime). Measures are sorted ascending by cost per tonne to build the curve.
One signed saving field. Annual net saving is a single signed figure combining energy savings, operating-cost change, and extra running cost — positive saves, negative costs. There is no separate operating-cost input.
Discount rate. The 3.5% default is the HM Treasury Green Book social discount rate — an engine constant cited to HM Treasury, not a MasterBrain value. The Private basis uses an editable 8% cost-of-capital preset. The rate materially changes every measure’s cost; a curve is only comparable to another built on the same basis.
Abatement sourcing. Direct abatement is user-entered tonnes; derived abatement is an avoided activity multiplied by an Ember grid or DEFRA fuel factor read live from MasterBrain and cited on the result. All measure economics are user-supplied — the tool holds no measure prices and fabricates none.
Carbon-price line. The threshold is user-set; in euro mode it can be seeded from the latest EU ETS allowance average as an editable default. The line defines the economically viable set but implies no recommended price.
Currency and display. Currency is a label only — all arithmetic is single-currency with no FX conversion. Outputs use magnitude-aware rounding; abatement is in tCO₂e, auto-scaling to ktCO₂e above 10,000; negative money values carry a minus sign. Per-measure figures are display-rounded, so a visible column may sum to within about £1 of the portfolio total, which is computed from unrounded intermediates.
Graceful handling. A derived measure whose factor cannot be resolved is excluded rather than defaulted; a measure with zero or negative abatement is excluded before any division. The tool never fabricates a value and never fails on missing data.
What a MACC is not. It is a static snapshot of measures treated as independent. It does not capture cost declines over time, interactions between measures, or the organisational barriers that keep economically favourable measures undone. Treat it as a structured input to a sequencing decision, not a fixed programme. Results are estimates for internal decision-support; the full methodological detail is on the paired marginal abatement cost methodology page.