A tunnel or in-bay automatic car wash concentrates hot water demand into a few peak hours. A gas boiler sized to that peak idles most of the year; an air-to-water heat pump sized the same way draws a demand charge every month for capacity used a few hours a week. The engineering approach is to size the heat generator to the daily average load and let a stratified thermal storage array carry the difference during the peak.
Demand is near zero overnight and concentrated into Saturday afternoon hours after a rain or road-salt event, when the conveyor runs at its rated cars per hour for as long as the queue holds.
This article works a stated tunnel design in US units, shows where the published European heat pump study transfers, and places the result against cold-ambient derate, demand charges, and water reclaim. It does not argue that every wash should electrify.
In this article
- Car washes are a measured water user and an unmeasured hot water user
- How a car wash heat pump water heater pairs with storage
- The peak hour, not the daily average, sets the storage volume
- Cold weather removes output exactly when the queue forms
- Water reclaim changes the heated volume before the heat pump sees it
- Where the economics land
- Regulatory and market timing, as of October 2026
- Seven checks before you specify anything
- Frequently Asked Questions
- Size the heat pump to the average and the storage to the Saturday
Car washes are a measured water user and an unmeasured hot water user
EPA's WaterSense at Work section on vehicle washes, revised November 2023, reports fresh water use of 44.8 gallons per vehicle for in-bay automatic washes and 30 gallons per vehicle for conveyor washes in the 2018 International Carwash Association study. In-bay figures range from 74.1 gallons without reclamation to 19.1 gallons with it. Reclamation systems cut fresh water use by 50 to 83 percent.
Hot water use per vehicle is not documented at the same standard. The EPA section contains no hot water or temperature figures. Natural Resources Canada, cited in the Capital Regional District fact sheet on vehicle wash water, estimated roughly 260 litres, about 69 gallons, of heated water per vehicle and about 1,300 gigajoules per year of water heating energy for an average commercial wash. That figure predates modern reclaim and is an upper bound rather than a design input.
The only engineering study of heat pump replacement for car wash boilers is European. Talaba, Serban, Hiris, and Balan, published in E3S Web of Conferences in 2025, report hot water use of 5 to 8 litres per minute per wash station at 55 °C from a 10 °C inlet, gas boilers of 35 kW for one or two stations rising to 90 kW for six, and a 200-litre buffer tank. Converted, 5 to 8 L/min is 1.3 to 2.1 gallons per minute per station, and the 45 °C rise is 81 °F. The study's euro-denominated payback figures do not transfer; its flow rates and heat pump performance curves do.
A trade analysis in Professional Carwashing & Detailing puts total water for a 120-car-per-hour friction tunnel at 53.85 gallons per car and does not separate heated from unheated water. Heated gallons per car depend on which arches and presoak applications are plumbed to hot water. The only reliable design number is a metered one.
How a car wash heat pump water heater pairs with storage
The heat pump is an air-to-water unit that lifts outdoor air heat into a hydronic loop at 55 to 60 °C, or 131 to 140 °F. The storage module holds a closed hydronic charge at that temperature. Wash water does not enter the vessel; it passes through a dual-coil heat exchanger of 32 mm DN corrugated stainless steel tubing, 40 to 80 metres long, with all connections through the top head plate. The stored volume is a hydronic buffer rather than a potable tank, which lets the vessel be charged above the delivery setpoint.
Stratified withdrawal governs discharge: hot charge enters at the top, and the coolest water is drawn from the bottom toward the heat pump. The top stays near charge temperature until late in the discharge.
The 700-gallon module loses roughly 2.4 °F per 24 hours at rest, which matters for a wash that charges overnight and discharges the following afternoon. An immersive electrode element rated 10 kW per tank provides backup heat, and a Modbus register interface lets the wash controller set the charge window. The thermal tank specification page lists the module sizes.
The options below compare hot water systems for a tunnel or in-bay automatic wash. The heat pump with modular storage is the subject of this article. Electric resistance is the default offered when gas is unavailable, and its demand charge exposure is roughly three times the heat pump case.
- Gas boiler or commercial gas storage water heater. Sized to peak cars per hour. Cold-ambient behavior: output constant; combustion efficiency unchanged. Demand charge exposure: none on the thermal side; gas commodity and fixed charges only.
- Air-to-water heat pump, no storage. Sized to peak cars per hour. Cold-ambient behavior: output falls with ambient; a unit sized to peak at 20 °C is undersized at -10 °C. Demand charge exposure: full connected electrical kW billed every month, used a few hours per week.
- Air-to-water heat pump with modular thermal storage. Sized to daily average load. Cold-ambient behavior: storage covers derate for one charge cycle; a backup element or extended run covers more. Demand charge exposure: connected electrical kW roughly one-third to one-half of the no-storage case.
- Electric resistance water heater. Sized to peak cars per hour. Cold-ambient behavior: output constant. Demand charge exposure: highest of the four; peak-hour electrical kW equals thermal kW.
The peak hour, not the daily average, sets the storage volume
Every design input below is an assumption to be replaced with the site's metered data. Peak throughput is 60 cars per hour. Heated water is 10 gallons per car, a labeled assumption—not a published benchmark—representing the presoak and wash arches plumbed to hot water, inside EPA's 30 gallons per vehicle of fresh water for a conveyor wash.
The assumed temperature rise is 70 °F, from a 50 °F winter inlet to a 120 °F wash setpoint. The European study's 81 °F rise is the upper end of what a US tunnel specifies.
Peak-hour heated water V_peak = cars/hour × heated gal/car V_peak = 60 cars/hour × 10 gal/car V_peak = 600 gal per hour
Peak-hour thermal energy Q_peak = V × 8.34 lb/gal × 1 Btu/lb-°F × ΔT Q_peak = 600 gal/hour × 8.34 lb/gal × 1 Btu/lb-°F × 70 °F Q_peak = 350,280 Btu per hour Q_peak = 102.7 kWh thermal per hour at a 70 °F rise, using 3,412 Btu per kWh thermal
A heat pump sized to that hour is a 103 kW thermal unit, roughly 350,000 Btu/h. The alternative is to size to the operating-day average. Assume a Saturday of 400 cars across a 12-hour operating day: 180 cars in the three-hour peak and 220 across the other nine hours.
Operating-day average thermal load Q_avg = (daily cars × heated gal/car × 8.34 lb/gal × 1 Btu/lb-°F × ΔT) / operating hours Q_avg = (400 cars × 10 gal/car × 8.34 lb/gal × 1 Btu/lb-°F × 70 °F) / 12 h Q_avg = 2,335,200 Btu / 12 h Q_avg = 194,600 Btu per hour Q_avg = 57.0 kWh thermal per hour at a 70 °F rise
Storage required to ride the three-hour peak E_storage = (Q_peak − Q_avg) × t_peak E_storage = (102.7 − 57.0) kWh thermal per hour × 3 h E_storage = 137.1 kWh thermal for the load at a 70 °F water-heating rise; module sizing below assumes a 35 °F usable storage delta-T
The 700-gallon module is rated at 108 kWh thermal at a 63 °F (35 °C) delta-T. A car wash cannot use the full nameplate delta because the discharge floor is the lowest tank temperature that still meets the wash setpoint through the coil. The usable delta-T is set here at 35 °F, a labeled assumption based on a 150 °F charge and a 115 °F floor.
The 150 °F charge temperature remains an equipment-selection assumption. Many R290 units top out at 55 to 60 °C, or 131 to 140 °F; R744 units reach higher. Confirm the charge temperature of the heat pump paired with the module and adjust the base case if 150 °F is not achievable. The discharge floor must still meet the wash setpoint through the coil.
Usable energy per 700-gallon module Q_module = volume × 8.34 lb/gal × 1 Btu/lb-°F × usable ΔT Q_module = 700 gal × 8.34 lb/gal × 1 Btu/lb-°F × 35 °F Q_module = 204,330 Btu Q_module = 59.9 kWh thermal at a 35 °F usable delta-T
Module count N = E_storage / Q_module N = 137.1 kWh thermal required for the 70 °F-rise load / 59.9 kWh thermal per module at a 35 °F usable delta-T N = 2.3 → 3 modules Three modules total 2,100 gallons and approximately 180 kWh thermal usable at a 35 °F delta-T.
At a 20 °F usable delta-T, each module holds 34.2 kWh thermal and the count rises to five. At a 50 °F usable delta-T, each holds 85.5 kWh thermal and two suffice. The delta is set by the wash setpoint and the heat pump's charge temperature, not by the tank.
The general method is set out in how to size thermal storage tanks for peak hot water demand. The thermal storage tank sizing calculator runs the arithmetic for other throughputs.
Thermal kWh is not electrical kWh. The conversion is the heat pump's COP at the ambient and lift of the design condition. The COP of 3.0 used below is an assumption, roughly the European study's R290 figure near 0 °C, and must be replaced with the site's design-condition data.
Electrical draw, heat pump sized to average P_elec = Q_avg / COP P_elec = 57.0 kWh thermal per hour / 3.0 P_elec = 19.0 kW electrical
Electrical draw, heat pump sized to peak without storage P_elec = Q_peak / COP P_elec = 102.7 kWh thermal per hour / 3.0 P_elec = 34.2 kW electrical
Avoided billed demand ΔP = P_peak − P_avg ΔP = 34.2 kW electrical − 19.0 kW electrical ΔP = 15.2 kW electrical
Annual demand charge avoided at an assumed rate S = ΔP × demand rate × billed months S = 15.2 kW electrical × $15/kW-month × 12 months/year S = $2,736 per year
The $15 per kW-month demand rate is an assumption that exposes the calculation structure, not a site tariff. NYSERDA's guide on storage for commercial users reports that demand charges make up 30 to 70 percent of many commercial bills and illustrates a Con Edison tariff with time-of-day demand rates of $8.32, $17.84, and $17.06 per kW. Peak demand charges explained covers interval metering and ratchets.
Each filled 700-gallon module weighs 6,046 lb on a 60-inch footprint and imposes 308 lb per square foot. The equipment room slab or floor requires structural review. Each module is 88.6 inches tall assembled, ships in panels, and is assembled through a standard doorway.
Cold weather removes output exactly when the queue forms
The European study publishes the derate curve a US designer needs. Its 20 kW nominal R290 unit delivers 22.9 kW thermal at 20 °C ambient, 17.1 kW thermal at 0 °C, and 11.6 kW thermal at -20 °C—roughly half its warm-weather output. Its COP falls from 4.0 at 20 °C to 3.1 at 0 °C and 1.8 at -20 °C.
The R744 (CO2) unit in the same study holds 60 °C output down to -25 °C, with a COP of 2.1 at -20 °C and 3.6 at 0 °C. DOE's heat pump water heater performance material from the Florida Solar Energy Center reports that compressor operation is generally limited below 50 °F ambient for the integrated residential units it tested. A car wash uses a dedicated low-ambient air-to-water unit.
The derate coincides with the load. In the northern two-thirds of the United States, the heaviest wash days follow snow and road salt, when ambient is below freezing and inlet water is coldest.
The European study's 200-litre tank with one station at 5 L/min runs 40 minutes without any heat source. With a 20 kW R290 unit at -25 °C, its upper layer held about 54 °C through a 60-minute test. The authors note that continuous sub-zero operation beyond roughly one to three hours requires an auxiliary heat source.
Storage is a daily-cycle asset. The three-module array carries one Saturday peak; it does not carry a three-day cold snap unless the heat pump keeps charging between peaks or the 10 kW per tank electrode element runs.
Three modules provide 30 kW of resistance backup, delivering 30 kWh thermal per hour at a COP of 1.0. Electrical draw is thermal output divided by COP: 30 kW thermal / 1.0 = 30 kW electrical, billed as demand when it runs. Rate the heat pump at the site's winter design ambient. The element is backup rather than capacity.
Water reclaim changes the heated volume before the heat pump sees it
Reclaim reduces fresh water volume, and heated water is nearly always fresh. The heated fraction of total water therefore rises as reclaim ratios rise even if heated gallons per car are unchanged. EPA's 1.9 to 4.9 gallons of reclaimed water per gallon of fresh for conveyor washes mean a 30 gal/car fresh figure sits inside 87 to 177 gallons of total water applied.
Reclaimed water returning from the pit also carries some upstream heat where reclaim is plumbed to heated arches. Neither hot-water effect is in the published data, so heated gallons per car must come from a meter on the hot water supply.
Reclaim is also a code matter in some states. California AB 2230, chaptered in 2012, requires an in-bay or conveyor wash permitted and constructed after January 1, 2014 to recycle at least 60 percent of its wash and rinse water or use recycled water from a supplier for that share. It exempts self-service washes.
EPA's WaterSense section also records retrofit limits of 35 gallons per vehicle for conveyor and 40 gallons per vehicle for in-bay automatic systems, citing the Texas Water Development Board.
Where the economics land
The gas price, electricity price, boiler efficiency, COP, and demand rate used in the comparison are calculation assumptions. Replace them with the site's own tariffs and equipment data before using the result for a project decision.
- Fuel cost per delivered MMBtu. At an assumed $1.20 per therm and 80 percent boiler efficiency, gas delivers heat at $15.00 per MMBtu. Using EIA's preliminary 2025 US average commercial electricity price of 13.41 cents per electrical kWh as the electricity-price assumption, and an assumed seasonal COP of 3.0, a heat pump delivers heat at $13.10 per MMBtu; at an assumed COP of 3.5, $11.23 per MMBtu. The EIA figure is from the February 2026 Electric Power Monthly. The crossover is close enough that the site's own tariffs decide it.
- Demand charge. Storage reduces connected heat pump electrical kW by roughly 45 percent in the worked example, from 34.2 to 19.0 kW electrical. Thermal energy storage for demand charge reduction sets out the full tariff method.
- Capital. Three 700-gallon modules at the September 2026 list price of $5,798 each total $17,394 for the storage line item alone, excluding heat pump, piping, controls, and installation. Published cost per stored kWh thermal is $54 for the 700-gallon module at its 63 °F rated delta-T.
- The published analogue. The Laundromat Puerto Rico case study pairs a 700-gallon tank with a 16 kW-thermal R32 heat pump on a site whose demand grew from 1,085 to 1,600 gallons per day. Displacing propane at $3.75 per gallon, annual cost fell from $6,610 to $2,830 on a $21,945 project, a 5.8-year published payback. A car wash has a sharper peak and a lower fuel price, so the payback will differ; the sizing logic is the same.
- Incentives. Utility incentives and rebates for commercial thermal energy storage explains how programs are structured. The incentive finder lists programs by postal code. Eligibility is determined by the program.
Regulatory and market timing, as of October 2026
DOE's October 6, 2023 final rule amending energy conservation standards for commercial water heating equipment, 88 FR 69686, codified at 10 CFR 431.110, requires compliance for models manufactured on and after October 6, 2026.
In an enforcement policy statement issued May 5, 2026, DOE stated that it will not seek civil penalties for violations of those amended standards for models manufactured between October 6, 2026 and October 6, 2027. DOE cited a pending petition for certiorari at the US Supreme Court, No. 25-879, over the D.C. Circuit decision that upheld the rule.
The policy is not a final agency action, and DOE may modify it. Re-check the DOE policy and Supreme Court docket for current status. An operator replacing a gas-fired commercial storage water heater in this window should confirm with the distributor which standard the offered model is certified against.
EIA's preliminary 2025 average retail price for commercial electricity is 13.41 cents per electrical kWh nationally, from the February 2026 Electric Power Monthly. State averages vary widely around that figure. Car washes frequently sit on small commercial tariffs with demand components the sector average hides.
Seven checks before you specify anything
- Meter the hot water supply for at least one full week including a Saturday, recording gallons and inlet and outlet temperatures. Heated gallons per car is a site figure, not a published one.
- Record the peak-hour car count and daily count for the same week. Confirm which arches and presoak applications are plumbed to hot water.
- Set the wash setpoint and heat pump charge temperature. Derive the usable delta-T from those two figures before computing module count.
- Rate the heat pump at the site's winter design ambient using the manufacturer's published derate curve, not its nominal rating at 7 °C or 20 °C.
- Obtain the current electric tariff, including demand rate, ratchet, and any time-of-use windows. Run the demand math against it.
- Confirm the equipment room slab or floor against 308 lb per square foot per module and the delivery path against panelized assembly.
- Confirm the reclaim system's fresh water make-up and whether any reclaim is plumbed to heated arches. The commercial laundry hot water systems article covers the parallel high-throughput case.
Frequently Asked Questions
How much hot water does a car wash use per car?
There is no published US benchmark for heated water per car. EPA reports 30 gallons per vehicle of fresh water for conveyor washes and 44.8 gallons per vehicle for in-bay automatic washes, of which only the arches plumbed to hot water are heated. The European study reports 5 to 8 litres per minute per station at 55 °C. A one-week hot water supply meter run is the only reliable design input.
What temperature does a car wash heat its water to?
Wash setpoints commonly fall between 110 and 130 °F. The European study designs to 55 °C, or 131 °F, from a 10 °C inlet. The setpoint and storage charge temperature together set the usable delta-T, which decides the module count more than any other variable.
Can a heat pump replace a car wash gas boiler in a cold climate?
Yes, with a derated rating. The published R290 curve loses roughly half its output between 20 °C and -20 °C. The design sizes the heat pump at the winter design ambient, charges storage overnight, and carries a backup element for extended cold. A CO2 unit holds output further into negative ambient at a lower COP.
How much thermal storage does a tunnel car wash need?
In the worked example, a 60-car-per-hour tunnel heating an assumed 10 gallons per car through a 70 °F rise needs approximately 137 kWh thermal of storage to ride a three-hour peak with the heat pump sized to a 400-car operating-day average. At an assumed 35 °F usable storage delta-T, that is three 700-gallon modules. These design inputs are site assumptions, not published benchmarks.
Do demand charges make a heat pump more expensive than gas?
They can. A heat pump sized to the peak hour draws roughly 34 kW electrical in the example and is billed for that capacity every month. Sizing to the average with storage cuts the draw to roughly 19 kW electrical. Both figures use an assumed COP of 3.0. Whether the result beats gas depends on the site's demand rate, gas price, and the heat pump's seasonal COP.
Does water reclaim reduce hot water demand?
Not directly. Reclaim reduces fresh water volume, and heated water is nearly always fresh, so heated gallons per car can be unchanged while the heated share of total water rises. Where reclaim is plumbed to heated arches, some heat returns from the pit. Only a meter on the hot water supply resolves the question for a specific wash.
Size the heat pump to the average and the storage to the Saturday
Car wash water use per vehicle is measured and public; heated water per vehicle is not. Heat pump output falls with ambient on the days a wash is busiest, and a boiler sized to the peak hour idles for most of the year.
Storage resolves the mismatch: a heat pump sized to the operating-day average charges a stratified array through the quiet hours, and the array discharges through the three-hour peak at roughly half the connected electrical kW a peak-sized heat pump would draw.
The worked design holds only as well as its four inputs: heated gallons per car, temperature rise, usable delta-T, and COP at the winter design ambient. Each must come from the site rather than from this page. Use this information to verify with the project's engineering services that the proposed solution will satisfy the project goals.


