Two paybacks, and why they differ
Simple payback is the first question anyone asks about an efficiency upgrade: how many years until it has paid for itself? Divide the net cost by the annual saving and you have it. It is easy, it is comparable across measures, and it is the number quoted on every proposal — but it makes two silent assumptions. It assumes a dollar saved in year fifteen is worth the same as a dollar saved next year, and it assumes energy prices never change.
Discounted payback drops both assumptions. Each year's saving is grown at your escalation rate and then discounted back to today at your discount rate, and the payback is the year in which those present values accumulate to the net cost.
Whether that makes payback shorter or longer depends entirely on which rate is larger. This is the part people get backwards. If you expect energy prices to rise faster than your money grows elsewhere — escalation above the discount rate — savings outrun discounting and the discounted payback is shorter than the simple one. If your discount rate is the higher of the two, future savings are worth less than their face value and the discounted payback is longer. When the two rates are equal, every year's saving has the same present value S/(1+r), and the discounted payback is exactly the simple payback multiplied by (1+r).
The calculator reports both, along with the lifetime net saving, so you can see the whole picture rather than one convenient half of it. For the specific case of an insulation upgrade, where the annual saving has to be derived from R-values and degree days first, use the attic insulation payback calculator and bring its annual saving here.
Every input, and how to choose it honestly
Net cost is what you actually part with. Take the installed price and subtract only rebates and credits you are confident of receiving. A utility rebate with a filing deadline you might miss, or a tax credit that is non-refundable when you owe no tax, is not a certainty. Overstating rebates is the fastest way to a payback figure that never materialises.
The annual saving must be a saving, not a bill. It is the difference between what you would have spent and what you will spend, in year-one prices. Beware of a quotation that computes savings against an unrealistic baseline — replacing a failed boiler with a new one saves nothing against a working boiler, only against the alternative you would otherwise have bought.
Escalation is the rate at which your energy price rises, not general inflation. If you use a nominal discount rate — what your money actually earns in cash terms — then escalation must also be nominal. If you prefer real terms, subtract inflation from both. Mixing a real escalation with a nominal discount rate is the most common technical error in this calculation, and it always makes the upgrade look worse than it is.
The discount rate is what the money would otherwise do. If you are borrowing, use the interest rate on the loan. If you are spending savings, use what those savings earn. There is no universally correct figure: a homeowner comparing an upgrade against a 5% deposit account and one comparing it against paying down 18% credit card debt should reach different conclusions, and they should.
Life is how long it keeps saving. Insulation, glazing and air sealing last as long as the building — forty years or more. Heating and cooling equipment is fifteen to twenty. Lighting is measured in operating hours. Set the life to the measure, not to how long you plan to stay: if you sell, the remaining value is captured through the sale price, which the home improvement ROI calculator handles.
Worked example: a $6,000 upgrade with a $1,200 rebate
A heat pump water heater is quoted at $6,000 installed, with a $1,200 utility rebate. It should save $520 in the first year. You expect energy prices to rise 3% a year, you use a 5% discount rate, and you assume a 20-year life.
- Net cost. $6,000 − $1,200 = $4,800.
- Simple payback. $4,800 ÷ $520 = 9.23 years.
- First year's present value. $520 ÷ 1.05 = $495.24.
- The ratio between successive years. Each year's saving is 1.03 times the last and is discounted by one more factor of 1.05, so present values fall by 1.03 ÷ 1.05 = 0.98095 each year — a geometric series.
- Cumulative present value. After 10 years it is $4,548.67; after 11 years, $4,957.16.
- Discounted payback. The net cost of $4,800 falls between them: 10 + (4,800 − 4,548.67) ÷ (4,957.16 − 4,548.67) = 10 + 0.615 = 10.62 years.
- Lifetime net saving. The 20-year present value of savings is $8,301.57, so the net saving is 8,301.57 − 4,800 = $3,501.57.
- Return on investment. $3,501.57 ÷ $4,800 = 72.9% over the life.
- Saving in year 10. $520 × 1.03⁹ = 520 × 1.30477 = $678.48 in that year's dollars.
The discounted payback is longer than the simple payback here — 10.62 years against 9.23 — because the 5% discount rate outweighs the 3% escalation. Flip the two rates, so prices rise 5% while you discount at 3%, and the relationship reverses: savings then outgrow the discounting, and the discounted figure comes in below the simple one. Neither ordering is universal, which is precisely why the calculator reports both rather than one.
Which number should decide it
Use simple payback to screen, not to decide. It is a good rough filter — under five years, do it; over twenty-five, look elsewhere — and it is the only figure that is comparable across proposals from different companies, because it needs no assumptions. Its weakness is that it stops counting at the moment of payback, so it treats a measure that pays back in eight years and dies in year nine identically to one that pays back in eight years and runs for forty.
Use lifetime net saving to decide. That figure is the net present value of the whole investment, and it is the one that answers "am I better off?". A positive NPV means the upgrade beats putting the money wherever your discount rate came from. It rewards long-lived measures properly, which simple payback does not: insulation with a fifteen-year payback and a forty-year life usually has a far better NPV than equipment with an eight-year payback and a fifteen-year life.
Treat ROI as NPV expressed per dollar, useful when comparing measures of very different sizes. A $600 measure returning 80% and a $6,000 measure returning 40% are not the same decision if you only have $600 to spend.
Sensitivity matters more than precision. Run the calculation twice more: once with escalation at zero, and once with the annual saving at 70% of the quoted figure. Real savings routinely come in below projections, partly because of the rebound effect — people who insulate often also turn the thermostat up, converting part of the saving into comfort. If the decision survives both stress tests, the assumptions were not doing the work. If it flips on either, the answer is genuinely uncertain and you should say so rather than pretending otherwise.
Watch for measures that are due anyway. When equipment has failed and must be replaced, the correct comparison is not efficient-versus-nothing but efficient-versus-standard. The cost to enter here is only the additional cost of the better option, and on that basis efficiency upgrades often look dramatically better than they do against a do-nothing baseline that was never available.
Typical measure lives, and what they do to the comparison
| Measure | Typical life | Why the life matters here |
|---|---|---|
| Air sealing | 10–20 yr | Cheap and fast-paying; short payback and modest lifetime value |
| Attic insulation | 40+ yr | Long payback often beaten by very long life — NPV tells a different story from payback |
| Replacement windows | 25–40 yr | Long life does not rescue a payback measured in many decades |
| Gas furnace or boiler | 15–20 yr | Payback must land well inside the life to be meaningful |
| Air-source heat pump | 15–20 yr | Saving depends on the fuel it displaces as much as on its own efficiency |
| Heat pump water heater | 10–15 yr | Shorter life makes rebates a large share of the result |
| LED lighting | 10–20 yr | Payback usually under two years; the analysis barely matters |
| Rooftop solar | 25–30 yr | Export tariffs and degradation need separate modelling, not a flat annual saving |
Lives are typical planning figures for residential work, not warranties. Where a measure's life is shorter than its payback, no discount rate makes it worthwhile.
Ways a payback figure gets flattered
- Counting a rebate you have not secured. Deadlines, income caps, contractor certification requirements and non-refundable credits all turn a headline incentive into a smaller one.
- Comparing against a baseline that no longer exists. If the old equipment has failed, the honest comparison is against the cheapest replacement, so only the incremental cost belongs in the calculation.
- An escalation rate chosen to make the answer work. At 8% a year, savings double in nine years and any measure looks good. Use a rate you would defend to somebody sceptical.
- Mixing real and nominal rates. Escalation and discount rate must be on the same basis. Mixing them silently biases the result, usually by two to three percentage points.
- Ignoring maintenance. Equipment measures carry servicing costs that fabric measures do not; a saving quoted gross of maintenance overstates the benefit.
- Ignoring the rebound effect. Households that improve efficiency often take part of the benefit as extra comfort rather than as a lower bill. That is a legitimate choice, but it reduces the cash saving the payback depends on.
- Assuming today's saving persists after a fuel switch. A measure that saves gas is worth much less once the house heats with a heat pump, and vice versa.
Where payback sits among the other ways to judge a measure
Payback is one member of a family of investment metrics, and knowing what the others do stops you asking payback to answer a question it cannot.
Net present value — reported here as the lifetime net saving — is the theoretically correct measure for a single accept-or-reject decision. Positive means do it.
Internal rate of return is the discount rate at which NPV falls to zero, which lets you compare an upgrade directly against an investment return. It is a good sanity check: an efficiency measure with an IRR of 12% is competitive with most things a household can do with the money.
Savings-to-investment ratio is the present value of savings divided by the net cost, and it is the standard metric in federal life-cycle cost analysis for buildings. A ratio above 1.0 means the measure pays for itself in present-value terms — the same test as a positive NPV, expressed as a multiple.
Two practical cautions. First, none of these metrics captures non-financial value, and for home energy work that value is real: a warmer house, a quieter one, fewer draughts, better resilience during an outage, and lower emissions. Many measures that fail a strict financial test are still worth doing, and the honest way to say so is to name the reason rather than to bend the discount rate until the numbers agree.
Second, sequence matters. Air sealing and insulation reduce the load a heating system has to meet, so doing them first lets you buy smaller and cheaper equipment afterwards. Run the fabric measures through this calculator before the equipment ones, and re-estimate the equipment saving on the improved house rather than the current one — otherwise you will double-count the same energy in two proposals. The cost side of all of it belongs in the renovation budget, and where the work is disruptive enough to keep a room out of use, the downtime cost calculator prices that separately.