After the energy shock: prioritising hotel retrofit investments

Retrofit options fall broadly into three categories, which behave differently enough to be assessed separately before being ranked together.

Operational controls are changes to how plant and systems are run: scheduling, setpoints, control logic, sequencing and the elimination of simultaneous heating and cooling. They cost little and can often be implemented quickly, but their benefit depends on the systems having the sensing and control capability to support them, and on consistent operation — without monitoring and accountability the saving can quietly decay.

Maintenance interventions address condition: fouled heat exchangers, leaking valves, failed insulation, unbalanced systems, filters and belts. A system that is not maintained will not perform to its design intent, however sophisticated its controls. This category is frequently under-invested because it appears in operating budgets rather than capital plans.

Capital works replace or add equipment: glazing, insulation, heat recovery, more efficient chillers or boilers, pumps, motors, lighting and controls upgrades — the largest investment, the longest service life and the most disruption during installation.

The ordering matters. Operational and maintenance measures lower the baseline consumption against which a capital measure is justified. Specify the capital project first, and it may be sized against a load that better operation would have reduced.

Simple payback divides the capital cost of a measure by its estimated annual saving, and expresses the result as the number of years required to recover the outlay. It is quick, transparent and widely understood — and routinely used as though it were a decision rule.

Its limitations are structural rather than incidental. It ignores everything after the payback point, so a measure with a long service life and a modest annual saving is ranked below one with a short payback and a short life. It takes no account of the time value of money. And it treats the annual saving as certain, which it is not: savings estimates are projections, and the realised figure depends on operation, weather, occupancy and tariff movement.

A discounted cash flow addresses the first two limitations by discounting future cash flows to present value over the measure's service life. It is more informative for larger capital decisions, and more demanding, because it requires a discount rate, a service-life assumption and a savings profile, each of which is itself a judgement.

The practical position is that both have a place. Simple payback is a defensible first screen for sorting a long list into a short one, and the short list should then be assessed on a discounted basis over service life, with the assumptions stated.

Consider a hypothetical measure with an assumed capital cost of 48,000 and an assumed annual saving of 12,000. Simple payback is 48,000 ÷ 12,000 = 4.0 years.

A second, hypothetical measure in the same property has an assumed capital cost of 15,000 and an assumed annual saving of 5,000, giving a payback of 15,000 ÷ 5,000 = 3.0 years. On simple payback the second ranks first, and on that narrow basis the ranking is correct.

Now suppose the first measure has an assumed service life of 20 years and the second an assumed service life of 5 years. Over a 20-year horizon, ignoring discounting, the first returns approximately 20 × 12,000 = 240,000 against a cost of 48,000, while the second would need replacing four times at approximately 15,000 each, returning 20 × 5,000 = 100,000 against a cost of about 60,000. The measure that ranked second on payback delivers more over the life of the asset. The illustration is deliberately simplified, assumes constant savings and tariffs, and is built from hypothetical figures.

The step that converts a projection into evidence is a verification plan, agreed before works begin, establishing what will be metered, over what period, against what baseline, and with what allowance for the variables not being controlled.

Verification is where occupancy and weather make the exercise harder. A saving measured across a period of different occupancy or weather from the baseline is not a clean comparison, and a credible plan accounts for that rather than presenting a raw before-and-after difference, using the property's own interval data and occupancy records.

The discipline has a further benefit: a measure that is verified, and found not to be delivering, can be investigated and corrected while it is still worth correcting. A measure assumed to be working continues to be counted as a saving long after it has stopped being one.