Payback, ROI, NPV and IRR for HVAC Efficiency Projects

Why the metric matters

Two HVAC projects with the same price can look very different depending on which yardstick you use. A finance team that screens on payback alone may reject a project that creates more value over its life than one it approves. Knowing what each metric measures, and what it leaves out, lets you present a project in the terms your decision-makers use while still showing its full value.

The four common metrics

Simple payback

Simple payback (years) = net project cost ÷ annual savings

Net cost is installed cost minus any incentives received. Payback is easy to explain and captures liquidity risk: the sooner the money comes back, the less exposure. Its weaknesses are that it ignores the time value of money and ignores every dollar of savings after the payback date. A three-year payback on a measure that lasts four years and a three-year payback on a measure that lasts fifteen years look identical.

Return on investment (ROI)

ROI has several definitions in practice, so state the one you use. A common lifetime version is:

Lifetime ROI = (total savings over the analysis period − net cost) ÷ net cost

An annual version divides one year's savings by the cost. ROI accounts for savings after payback but, in its simple form, treats a dollar saved in year ten the same as a dollar saved today.

Net present value (NPV)

NPV = −net cost + Σ (savings in year t ÷ (1 + r)^t), summed over the analysis period, where r is the discount rate and ^t means raised to the power of the year number t.

NPV converts every future dollar of savings into today's dollars and subtracts the investment. A positive NPV means the project earns more than the discount rate. It is the most complete single measure because it captures both the size and the timing of the savings.

Internal rate of return (IRR)

IRR is the discount rate at which NPV equals zero. It reads like an interest rate, which makes it easy to compare with a hurdle rate or a cost of capital. It can mislead when comparing projects of very different size or life, because it says nothing about how many dollars a project creates.

Savings-to-investment ratio (SIR)

Federal life-cycle cost analysis also uses SIR: the present value of savings divided by the present value of the investment. An SIR above 1.0 means the project is cost-effective at the chosen discount rate. SIR is handy for ranking several candidate projects against a fixed budget.

Worked example: two projects, two rankings

Illustration only. "Measure A" and "Measure B" are hypothetical efficiency projects, not quotes or results for any product.

How the NPV for Measure A at 8% is built: the present value of $12,000 a year for 10 years at 8% is $12,000 × 6.7101 (the 10-year annuity factor) = $80,521. Subtract the $40,000 cost and the NPV is $40,521.

In this case all metrics favor Measure A. But notice how the gap changes with the discount rate. At 3%, Measure B's NPV nearly triples, because its savings run longer and a low rate gives distant years more weight. With a different pairing, say a project with a long life and a moderate payback against one with a short life and a fast payback, payback and NPV can point in opposite directions. When they do, NPV is the better guide to value, and payback tells you about risk and cash timing.

Choosing a discount rate

The discount rate expresses what money is worth to your organization over time. Common choices:

• Corporate hurdle rate or weighted average cost of capital. Typical for private owners; ask your finance team.
• Borrowing rate. If the project is financed, the interest rate is a floor.
• Federal rates. For federal energy and water conservation projects, the 2025 Annual Supplement to NIST Handbook 135 sets a DOE real discount rate of 3.0% (a floor that applied because the calculated value was lower) and a nominal rate of 4.5%. These rates apply until NIST publishes the next edition; check for an update before you use them.

Be consistent. A real rate excludes inflation and pairs with savings in constant dollars. A nominal rate includes inflation and pairs with savings that rise with expected energy prices. Mixing a nominal discount rate with flat, constant-dollar savings understates value; mixing a real rate with escalating savings overstates it. NIST's annual supplement publishes energy price escalation indices for this purpose, built from EIA's Annual Energy Outlook projections.

Energy price escalation

If you expect electricity prices to rise, the dollar value of each kWh saved rises too. In the example above, letting Measure A's savings grow 2% a year (a nominal assumption) raises its NPV at 8% from $40,521 to about $47,074. Use escalation only with a nominal discount rate, and show results with and without it so reviewers can see the effect.

Setting the analysis period

Do not analyze savings beyond the life of the equipment they depend on. If a measure is applied to a 15-year-old rooftop unit that will likely be replaced in five years, the analysis period is five years, even if the measure itself could last longer. For efficiency treatments applied to existing equipment, the remaining life of that equipment is the natural limit. CryogenX4, for example, states that its one-time treatment is intended to last for the remaining life of the equipment, so a sound analysis uses the equipment's expected remaining life as the period. For help estimating remaining life, see repair, retrofit or replace.

Incentives, demand savings and maintenance effects

• Incentives. Utility incentives typically arrive after installation and verification, sometimes months later. Model them as a cash inflow in the year you expect to receive them, not as a reduction in day-one cost. See Texas utility commercial incentives.
• Demand savings. Include them only where your tariff supports them, and list them separately from energy savings. See how to calculate HVAC energy savings.
• Maintenance and avoided costs. If a project changes maintenance costs or defers a capital replacement, those cash flows belong in the analysis too. Deferral is covered in capex vs. opex HVAC decisions.

Stress-test the savings estimate

The biggest risk in any HVAC efficiency analysis is not the discount rate; it is the savings figure. Run at least three cases. In the Measure A example, if savings come in at half the estimate ($6,000 a year), payback doubles to 6.7 years and NPV at 8% falls to about $260, essentially breakeven. A project that only works in the optimistic case is a candidate for a measured pilot before full rollout.

That is why performance contracts tie payments to measured savings. FEMP's M&V Guidelines Version 5.0, used for federal energy savings performance contracts, and the IPMVP framework describe how to verify savings so that the cash flows in an NPV model rest on data rather than assumptions. See how HVAC energy savings are measured.

When a payback threshold rejects a good project

Many organizations screen projects with a maximum payback, for example three or five years. A hard cutoff is simple to apply but can reject measures that last much longer than the threshold and create substantial value over their lives. If your threshold rules out a project with a strong NPV, present both numbers and ask whether a longer threshold or an NPV test applies to efficiency measures. Some organizations also set aside a separate pool for efficiency projects, judged on NPV or SIR, so that they do not compete directly with projects judged on strategic grounds.

Presenting the numbers

A one-page summary that finance teams can act on usually includes:

1. Net cost, with incentives shown separately and timed realistically.
2. Annual energy savings (kWh and dollars) and demand savings, each with its basis.
3. Simple payback, NPV at the organization's discount rate, and IRR.
4. Low, expected and high cases for the savings estimate.
5. The analysis period and why it was chosen.
6. How savings will be verified after installation.

For a fuller template, see building the business case.

Where CryogenX4 fits

CryogenX4 reports a typical payback on its treatment of 12 to 36 months and energy savings of up to 30%, noting that results vary by equipment condition. Use those figures to decide whether a project is worth investigating, not as inputs to an approval model. A measured pilot on representative units gives you a savings figure for your own equipment that can go into the NPV, IRR and payback calculations above.

Next step

Ask your finance team for the discount rate and payback threshold they use, then enter your own savings estimate in the HVAC savings calculator and build the low, expected and high cases from the result.