A Hotel Decarbonization Cost Curve: Comparing Measures by Cost per Emissions Reduction

The question most hotel plans get wrong

When an owner or asset manager asks which green project to approve first, the usual answer ranks measures by total capital cost or by the largest advertised percentage saving. Both answers are misleading. A low-cost measure can remove almost no carbon, while a costly retrofit can be the cheapest way to avoid a tonne if it abates many tonnes over a long life.

A 200-room business hotel in a cold climate, a 40-room seaside resort and a heritage riad guesthouse face different energy bills, occupancy curves, weather and utility prices. A ranking borrowed from another property almost never transfers. What transfers is a method: compare every candidate by the cost of removing one tonne of carbon dioxide equivalent over the lifetime of the investment. That is the logic of the marginal abatement cost curve, a tool long used in energy policy and now applied to hotel real estate.

Most owners also forget that carbon accounting separates on-site fuel (Scope 1, such as gas boilers and the fleet), purchased electricity (Scope 2) and everything bought in, from laundry to food (Scope 3). Measures in each scope sit at very different points on the curve, so the inventory must come before the ranking.

• Cost per tonne of CO2 abated is the metric, not project cost or a marketing percentage
• A cheap fix that saves 5 tonnes can rank below an expensive retrofit that avoids 1,000 tonnes
• The boundary and emission scopes must be fixed before measures can be compared fairly

Cost per tonne: the metric that makes measures comparable

Marginal abatement cost is the net present cost of a measure over its lifetime divided by the tonnes of CO2e it abates over that same lifetime. Net cost includes the incremental capital outlay minus the operating savings from reduced fuel and electricity, discounted over the equipment life. When savings exceed costs the number turns negative: the measure makes money while cutting emissions. Textbook treatments note that in the classic McKinsey curve, LED lighting sits at roughly negative cost because it saves more electricity than it costs.

The denominator matters as much as the numerator. Two measures with identical capex can have completely different cost per tonne if one lives five years and the other twenty, or if one displaces gas and the other trims lighting at low occupancy. This is why a simple payback figure is insufficient: payback ignores the full lifetime and the quantity of carbon actually displaced.

Capital intensity is a separate, complementary view: it divides additional upfront investment by lifetime emissions saved and shows which technologies demand the most cash even when they are financially attractive. A hotel should read both, because a negative-cost but capital-hungry project can still strain the budget if sequenced badly.

• Formula: cost per tonne = lifetime net present cost divided by lifetime tonnes abated
• Negative cost means the measure pays for itself while removing carbon
• Capital intensity (upfront cash per lifetime tonne) is read alongside cost per tonne

What typically sits on a hotel's cost curve

When dozens of hotel measures are ranked by cost per tonne, they sort into a recognizable shape. At the left edge sit fast operational and lighting measures that are negative or near-negative cost: LED replacement, occupancy and key-card guest-room controls, heating and cooling setbacks in vacant rooms and wings, laundry and kitchen scheduling, and boiler and chiller tuning. These are the quick wins that should never wait for a climate budget.

The middle of the curve holds heat recovery on ventilation and laundry, solar thermal or heat-pump hot water, and air-source heat pumps replacing gas space heating. Their position moves a lot: where electricity is cheap, gas is expensive and the grid is clean, electrification jumps left toward negative cost; where gas is cheap and the grid is dirty, it slides right. Rooftop photovoltaic systems sit in a similar price- and climate-sensitive band, with payback that depends on local tariffs, export rules and sun.

The expensive tail contains deep envelope insulation behind single-glazed heritage facades, ground-source heat pump fields, full kitchen electrification, and on-site carbon capture for boilers, which remains costly per tonne and is best treated as a last resort or novelty rather than a backbone measure.

• Negative or near-negative band: LED, occupancy controls, HVAC setbacks, boiler and chiller tuning
• Middle band: heat recovery, heat-pump and solar hot water, air-source heat pumps
• Price-sensitive: rooftop PV and electrification shift with electricity tariffs and grid carbon
• Expensive tail: heritage-window retrofits, ground-source loops, on-site carbon capture

What real retrofit projects show

Documented whole-property retrofits show the practical shape of the curve. An EBRD-financed hotel in Kusadasi, Türkiye, spent about €285,000 on a variable-refrigerant-flow system, heat pumps, LED lighting and solar water heaters; it cut energy costs by roughly €60,000 a year (payback under five years), avoided about 436 tonnes of CO2 a year and reduced its carbon emissions around 65%.

A Moroccan project converting old riads in Marrakech combined floor and wall insulation, heat pumps, solar hot water, inverter air conditioning and 334 LED lamps for about €89,000. It cut energy consumption by about 66%, equal to roughly 150 MWh and 61 tonnes of CO2 a year, with a payback near two and a half years. Both cases show that a coordinated package earns more than the sum of isolated parts.

The Pembroke Kilkenny hotel in Ireland spent about €524,000 on a multi-year upgrade that included high-efficiency heat recovery with a CO2 sensor, induction hobs and more efficient appliances. The project cut about 107 tonnes of CO2 and 477,000 kWh of energy a year while saving about €36,000 annually in costs, supported by roughly €157,000 in grants. The grants matter: they shorten payback and pull measures leftward on the curve, a reminder to price available subsidies and utility incentives into every line.

• Türkiye hotel: ~€285k package, ~€60k annual energy saving, ~436 tCO2/yr, ~65% cut
• Marrakech riads: ~€89k package, ~66% energy cut, ~61 tCO2/yr, ~2.5-year payback
• Kilkenny hotel: ~€524k upgrade with grants, ~107 tCO2/yr and ~€36k annual saving
• Grants and utility programs shift measures left on the curve and should be included in the math

How to build your own curve from your data

Start with a baseline that cannot be argued with: twelve months of utility data per building, ideally broken out by meter, plus occupancy records and floor area. Convert these into intensity metrics such as kilowatt-hours and kilograms of CO2 per occupied room or per square metre. Benchmarking initiatives such as the Cornell Hotel Sustainability Benchmarking Index let an owner compare those figures against a peer set in the same city, asset class or climate zone, and tell whether the property is a normal, efficient or wasteful case.

Next, list every candidate measure with three figures each: estimated annual energy saving, expected lifetime in years, and incremental capital cost. The Cornell-style benchmark work shows that average carbon per stay has fallen over time and that energy intensity per square metre varies widely by segment and region, which is exactly why a generic measure list will not do; the owner's own benchmark decides how much headroom exists.

For each measure, compute lifetime net present value of cost (capital minus saved energy, discounted), estimate lifetime tonnes abated using local emission factors for the electricity mix and gas, and divide. Rank the results from most negative to most expensive and draw the staircase curve. Then decide the operating principle: approve everything at or below zero cost first, then work up the curve until the budget or the carbon target is met, using the internal carbon price or subsidies to justify items above zero.

Rebuild the curve annually with measured, not promised, savings. The multi-year CHSB studies found headline decreases in energy and emissions per square metre, but also warned that occupancy swings and pandemic closures distorted single-year results, so a one-off year should never set policy.

• Gather 12 months of metered energy, occupancy and floor-area data as the baseline
• Convert to intensity metrics and compare to Cornell Hotel Sustainability Benchmarking Index peers
• Score each measure: annual saving, lifetime, incremental capital, then NPV and cost per tonne
• Rank from negative to expensive; approve zero-and-below items first, then fund up the curve
• Refresh yearly with measured savings, not vendor promises

Limits: why your curve is a tool, not a verdict

The cost curve relies on assumptions that age quickly. Static prices from the day it is built tend to overstate the cost of fast-falling technologies such as solar; energy tariffs, carbon prices and grid emission factors change year to year, so a heat pump that is marginal today can become clearly negative cost tomorrow. Treat the curve as a live planning document, not a one-time spreadsheet.

Data quality is the other soft spot. Benchmarks built on self-reported data, meter faults or wrong floor-area totals produce a curve that rewards accounting errors. The same caution applies to carbon accounting rules, which differ across schemes; a tonne avoided on site is not the same as a tonne offset by a credit, and the two should never be mixed in the same ranking.

Finally, the curve shows economics, not feasibility. Heritage constraints, grid capacity, fire and ventilation codes, and guest-experience risk can veto the cheapest line. For financing, tax treatment and grant eligibility, consult qualified local advisers, because rules differ by jurisdiction. None of this weakens the core idea: if you cannot compare measures by cost per tonne, you cannot claim to have chosen the cheap carbon first.

• Static prices overstate the cost of solar and other falling-cost technologies; update inputs yearly
• Carbon accounting conventions and offsets must not be mixed into the same curve
• Feasibility (heritage, grid, codes, guest comfort) can veto the cheapest line
• Use local advisers for grants, tax and financing; rules differ by jurisdiction