Methodology & assumptions · v3.1

How the Water Heater Lifecycle Calculator works

Every equation, default and source behind the numbers, published so experts can check the math and anyone can argue with an assumption. If you think a number is wrong, say which one and why.

Purpose

Most water heater calculators compare one year of energy at one flat price. This one asks a different question: what does each option cost per year of hot water, over its whole life, in a specific home? That means climate-specific heat demand, the actual utility tariff, units that last different lengths of time, repairs the warranty won't cover, days without hot water, end-of-life replacement, and what happens when the grid goes down.

It was built for one real decision, a 55-gallon electric tank failing in a Miami garage, and generalized so others can use it.

1. Heat needed

Monthly heat delivered to the water, in kWh of heat:

Q_month = gal/day × 8.34 lb/gal × (T_setpoint − T_supply,month) ÷ 3,412.14 BTU/kWh × days_in_month

Hot water use defaults to 15 gallons per person per day (60 for four people). Standby losses are not modeled separately; they are carried inside each unit's rated UEF, which includes them under test conditions.

2. Climate and supply water

Monthly mean air temperatures come from climate normals for each preset city (Exuma values are approximate). Supply-water temperature uses the NREL Building America mains model (Burch & Christensen):

T_supply = T_air,annual + 6 + ratio × (ΔT_air ÷ 2) × sin(0.986 × (day − 15 − lag) − 90°) ratio = 0.4 + 0.01 × (T_air,annual − 44) lag = 35 − (T_air,annual − 44) days

Users can shift the result with a supply-water adjustment. The air around the unit depends on placement:

PlacementAir around the unit
GarageOutdoor mean + 5°F + 15% of any shortfall below 65°F
Garage, ducted to attic airOutdoor mean + 12°F
Inside, conditioned72°F (heating/cooling interaction not modeled)
Outdoor closetOutdoor mean
Why this matters: Miami tap water averages about 83°F against the 58°F assumed by the national efficiency test, so a Miami home needs about 40% less heat for the same hot water. National savings claims overstate Miami savings for that reason alone.

3. Equipment efficiency

Standard electric tanks and tankless units

kWh = Q ÷ UEF

Heat pump water heaters

COP_month = UEF × f_air(T_air) × f_set(T_setpoint) f_air = 1 + 0.015 × (T_air − 67.5), bounded 0.60 to 1.35 f_set = 1 − 0.014 × max(0, T_setpoint − 125), floor 0.50 kWh = Q × [ (1 − e) ÷ COP + e ÷ 0.98 ]

e is the element share: the fraction of heat delivered by the backup electric elements instead of the compressor. Defaults are 10% in heat-pump-only mode in a warm space, 8% for the inverter unit, 3% for the predictive unit, and 35% as the field-average for hybrid mode (NEEA end-use data). Rated UEF is measured at 67.5°F air, which is why warm garages raise effective efficiency.

4. Tariff and time of use

The default tariff is FPL RS-1, effective January 2026, per kWh before tax:

ComponentFirst 1,000 kWhAbove 1,000 kWh
Base energy7.865¢8.865¢
Fuel2.893¢3.893¢
Conservation, capacity, environmental, storm protection, transition credit1.500¢1.500¢
Total12.258¢14.258¢

The water heater's kWh are billed 75% at the upper tier by default (months the home exceeds 1,000 kWh), plus 12% for taxes and fees (gross receipts, franchise fee, municipal utility tax; this varies by city). The $10.52 customer charge doesn't change with the water heater and is excluded. Prices rise 3% a year by default.

Time of use (FPL RTR-1)

RTR-1 adds 14.636¢ on-peak and credits 6.255¢ off-peak on top of RS-1. On-peak hours are weekdays noon–9 pm April–October and 6–10 am plus 6–9 pm November–March. Weekends and holidays are off-peak.

usable storage (kWh) = gallons × 0.75 × 8.34 × (T_store − T_supply) ÷ 3,412.14 cover = min(1, usable storage ÷ (window share × daily heat)) on-peak fraction = window share × (1 − cover) × 5/7

T_store equals the setpoint, or 140°F for units with a mixing valve and predictive control. Default window shares (hot water drawn inside on-peak hours) are 40% in summer and 55% in winter. Shifting costs energy: an 18% summer penalty for resistance tanks (FSEC lab data), 5% for heat pumps without predictive control, none for predictive units. Heat pumps without predictive control get 8 more points of element share on time of use, reflecting field data that units lean on elements to recover after a shed window. Tankless heaters cannot shift.

5. Demand charges

Optional, in $ per kW-month; FPL RS-1 has none. Each unit's contribution to the billed monthly peak:

tankless 0.9 × rated kW standard tank 0.5 × 4.5 kW × (1 − 0.8 × shifted) heat pump [0.8 × compressor kW + 0.5 × 4.5 kW × min(1, 2e)] × (1 − 0.8 × shifted)

"Shifted" applies only when demand is measured on-peak and the home is on time of use.

6. Lifecycle and equal footing

Options with different lives are compared on equivalent annual cost, the standard method for assets with unequal lives. Inputs are today's prices. Two discount rates are used: the nominal rate d for electricity, whose prices escalate at their own rate, and the real rate r = (1 + d) ÷ (1 + inflation) − 1 for equipment and labor, which rise with inflation.

Timing conventions

Six non-overlapping cost categories, per unit life

Each value is a present value at the start of that unit's life, in today's dollars. a = annuity(r, L) ÷ (1 + inflation) is the present value of a cost of $1 a year at today's prices, paid in years 1 to L.

purchase & install = installed cost − rebate (+ $2,500 panel upgrade for the first tankless unit, if checked) maintenance = upkeep per year × a repairs = repairs per life × [parts × (1 − min(1, parts warranty ÷ L)) + labor × (1 − min(1, labor warranty ÷ L))] ÷ L × a downtime = repairs per life × days down per repair × cost per day ÷ L × a + days down at replacement × cost per day ÷ (1 + r)^L emergency premium = premium % × installed cost of the incoming unit ÷ (1 + r)^L

The replacement event is priced on the incoming unit: its price sets the emergency premium and its days down set the replacement downtime. If the user chooses to replace on a plan, the premium and the replacement downtime are zero. Days without hot water are counted only in the downtime category; repairs are parts and labor only.

From one life to the comparison period

each category's annual cost over that life = category value ÷ annuity(r, L) present value over the period = Σ over each unit life starting at year s and ending at year e: annual cost × [annuity(r, min(e, N)) − annuity(r, s)] electricity present value = Σ_{y=1..N} year-1 bill × (1 + escalation)^(y−1) ÷ (1 + d)^y cost per year = present value ÷ annuity(r, N) (N = years to compare)

A unit whose life runs past the end of the period contributes only the years inside it, which is the same as crediting its remaining life. Because every category is carried through the same steps, the chart's six pieces (five equipment categories plus electricity) add exactly to the total shown. Nothing is rescaled.

Chains and the deferred option

An ordinary option is a chain of identical units: bought at year 0, replaced at L, 2L, and so on. "Standard tank now, heat pump at next failure" is the tank for its life, then the chosen heat pump, then that heat pump's own replacements. The switch is priced on the heat pump: its installed cost (plus a panel upgrade if the incoming unit is tankless), its emergency premium and its days down.

7. Cost per year vs. cash

The calculator reports two different things on purpose.

MeasureWhat it answersIncludesLeaves out
Cost per year (equivalent annual cost) and present valueWhich option is cheaper over its whole life, compared fairlyAll six categories, discounted; credit for life left in the last unit at the end of the periodNothing in the six categories
Cumulative cash and break-even yearWhen the money actually movesPurchases, emergency premiums, upkeep, expected repairs and electricity, in nominal dollars in the year paidDays without hot water (a real cost, but not a payment); any credit for remaining life; a replacement that falls exactly in the final year (it belongs to the next period)

The two totals will not match, and should not. Cash can make a strategy look worse when it has just bought a new unit near the end of the period, or better when its next purchase falls just outside it. Use cost per year to choose; use cash to plan.

Break-even is the first year in which the candidate's cumulative cash is at or below the baseline's. Because a heat pump's replacement is a large single payment, cash can fall behind again for a year or two around it. The calculator says so, and gives the year from which the candidate stays ahead through the end of the period. In the Miami example the heat pump breaks even in year 6, falls behind in years 12–13 when it is replaced, and stays ahead from year 14.

8. The 2029 standard

DOE's 2024 consumer water heater standards require electric storage water heaters above 35 gallons made on or after May 6, 2029 to meet heat-pump efficiency (UEF 2.30 for 20–55 gallons). The "standard tank now, heat pump at next failure" option runs the standard tank for its life, then the chosen heat pump and its replacements for the rest of the period (section 6). Units made before the date may still be sold. DOE listed the rule as in force as of August 2026; check its status before relying on it.

9. Outages and CO₂

outage kWh/day = daily heat ÷ (UEF × 1.2) for heat pumps in heat-pump-only mode = daily heat ÷ UEF otherwise

A backup source must supply the unit's peak draw to run it at all: about 27 kW for whole-home electric tankless, 4.5 kW for a standard element, about 0.5 kW for a compressor. The calculator reports the share of a battery used per day, and, using the outage days per year, the yearly backup energy (battery) or yearly generator fuel cost. Outage costs are shown for decision-making but are not added to cost per year; the small amount of grid electricity not bought during outages is also ignored. For the deferred option the outage figures describe the unit installed today.

CO₂ over the comparison period = grid kWh × grid emission factor (lb/kWh) ÷ 2,000, in short tons. Grid kWh exclude the share met by your own solar. The deferred option counts the tank's kWh until the switch and the heat pump's after it. The emission factor is held constant.

10. Hypothetical mortgage-payment equivalence

Monthly all-in modeled savings × 150.3 illustrates an equivalent financed amount at 7% over 30 years. These savings include lifecycle costs and downtime, not just electricity payments. This is a hypothetical mortgage-payment equivalence, not an estimate of property value or a lender valuation.

11. Default assumptions

Installed costs are Miami estimates for a garage swap before rebates (FPL offers no heat pump water heater rebate; Florida's state rebates have not launched; the federal 25C credit ended with 2025). Replace them with local quotes. Each option has a "serviced locally" box. In the Miami example every option is ticked except Cala, which has no Miami installer or service network today; untick any unit without local service where you live. The headline winner is chosen among ticked options.

OptionInstalledUEFElement shareGalLifeUpkeep/yrWarranty parts/laborRepairs/lifeParts/labor per repairDays down repair/replace
A.O. Smith 55 gal electric (like-for-like)$1,9000.93—558$256 / 11$60 / $2001 / 1
Rheem RTEX-27 tankless$3,2000.99——15$801 / 12$250 / $2503 / 2
Bradford White AeroTherm G2 65$3,6004.2010%6512$4010 / 11.5$300 / $3000.5 / 2
A.O. Smith Voltex AL 66$3,8003.7010%6612$4010 / 11.5$300 / $3000.5 / 2
Rheem ProTerra 65$3,2004.0510%6512$4010 / 11.5$300 / $3000.5 / 2
LG inverter 58 (APHWC502L)$3,5004.208%5312$4010 / 11.5$300 / $3000.5 / 2
Cala 65$4,0003.613%6515$4010 / 31.5$300 / $3000.5 / 2

Other defaults: 120°F setpoint; $40 per day without hot water; 15% emergency premium; 5% discount rate; 2.5% equipment and labor inflation; 3% annual electricity price increase; 24-year comparison (a common multiple of 8 and 12); grid CO₂ 0.85 lb/kWh.

12. Validation

Wanted: metered data from real installations, especially Florida, to tighten element share, the air-temperature curve and the time-of-use penalties.

13. Input checks

The calculator refuses to compute on inputs that would produce meaningless results. Instead it shows a message, highlights the field and replaces results with "—". Selected checks:

14. Known limitations

Sources

Change log