When commercial drain flow and hot water draw are out of phase, recovered heat has to be stored before it is useful. The preheat tank is sized on that mismatch, not on the exchanger. A gravity-film exchanger without storage recovers useful heat only while drain and cold supply flow at the same time; batch laundry loads and a hotel’s afternoon-laundry-to-morning-shower mismatch require storage. The final heater still carries the lift from preheat temperature to delivery temperature.
Key Takeaways
Oak Ridge National Laboratory estimates that the equivalent of roughly 350 billion kWh of hot water goes down drains in the United States each year. CDC guidance states that laundries consume 50 to 75 percent of a hospital’s total hot water. The resource is substantial; its timing determines whether the building can use it.
In a residence, a shower drains while the same shower draws cold makeup, and a vertical exchanger on the drain stack recovers heat without storage. Commercial flows are typically out of phase. A washer-extractor dumps 200 gallons in three minutes and fills again ten minutes later. A hotel’s on-premise laundry runs all afternoon while the guest shower peak ends by 9 a.m. A dish machine drains in pulses between racks.
This article treats phase mismatch as the design variable. It compares three hardware classes, works the Btu balance for a stated laundry load, sizes the preheat tank at a stated usable ΔT, and converts recovered thermal energy to avoided heat pump input at a stated COP. It also distinguishes equal-flow from heater-only gravity-film plumbing and explains why stored preheat removes that flow-balance penalty.
In this article
Commercial drains carry more heat than the published residential cases
The residential evidence base is consistent and modest. An Oak Ridge National Laboratory field study of a Duluth, Minnesota triplex, reported at the ACEEE Hot Water Forum, measured 25 to 30 percent savings on total hot water energy from a 50 percent effective gravity-film exchanger.
Efficiency Vermont’s program guidance puts the upper bound for a well-placed residential unit at a 60 percent reduction in water heater energy. ORNL’s 2024 laboratory characterization of a drain exchanger for heat pump water heaters measured effectiveness between 45 and 85 percent across inlet temperatures of 9 to 36 C and flow rates of 0.03 to 0.28 kg/s. A designer should expect a range rather than a single rated figure.
The commercial cases in the same ACEEE presentation are larger and use storage. A Holiday Inn in St. Catharines, Ontario installed three multi-coil exchangers in its laundry drain trough in 2014, metered a 27 F average rise on incoming water and reported a 3.5-year payback. A textile mill in Exeter, Ontario pumped drain water into a holding tank so a vertical exchanger could run between wash cycles and reported a 2.7-year payback. These are named case studies reported in a manufacturer-authored conference presentation, not a peer-reviewed study.
A restaurant pilot run by the Energy Center of Wisconsin and reported by the Clean Energy Resource Teams recovered 26 percent of the hot water energy originally delivered, the best of four pilot sites, at retrofit costs of $1,500 to $3,800 for a gravity-film unit. Commercial kitchen hot water storage covers the hot side of the restaurant plant, which the pilot did not change.
The largest documented commercial installation uses a heat pump rather than a passive exchanger. EPRI’s 2021 report on a 60-unit multifamily building in North Vancouver documents a wastewater-source heat pump drawing on 2,000 gallons per day of drain water at an average 68 F exit temperature, running at an average COP of 3.7 and producing 100 percent of the building’s hot water at 140 F. The system held 878 gallons of wastewater in storage, which the report credits with allowing peak load shifting.
The scale of the resource follows from water use. EPA’s WaterSense at Work guidance puts washer-extractor consumption at 2 to 4 gallons per pound of laundry and tunnel washers at 2 gallons per pound or less. It attributes about 90 percent of a clothes washer’s energy to water heating. CDC guidance notes that at least 160 F for 25 minutes is commonly recommended for hot-water washing in healthcare laundries, so blended drain water arrives well above ground temperature.
Three hardware classes, and only one works without storage
The gravity-film exchanger, pumped drain-to-tank exchanger and wastewater-source heat pump recover the same heat by different mechanisms. The mechanism determines whether storage is optional.
The gravity-film unit is the residential default and the one described on DOE’s drain-water heat recovery page. DOE states that a non-storage unit provides useful heat only while cold water and drain water flow at the same time, and that a dishwasher or clothes washer requires a unit with storage capacity. That limitation governs the commercial laundry case.
The pumped drain-to-tank class is where the commercial case studies fit. The Holiday Inn exchangers sat in the drain trough, and the textile mill held drain water in a tank so the exchanger could run on its own schedule.
The wastewater-source heat pump replaces the passive exchanger with an evaporator and adds compressor work, so it reaches delivery temperature rather than a preheat temperature. Air source vs water source heat pumps for domestic hot water covers that source selection, including wastewater.
Where the recovered heat goes: the coil tank as the preheat vessel
A preheat tank receives heat from the drain side whenever drain water is flowing and releases heat to the potable supply whenever hot water is drawn. The two events do not have to coincide.
TEHQ’s modular thermal tank serves this role with a dual-coil heat exchanger: a loop of 32 mm DN corrugated stainless steel tubing, 40 to 80 metres, with all connections entering and leaving through the top head plate. The drain-side loop charges one coil. The potable supply passes through the other coil on its way to the final heater.
The stored volume between the coils is a closed hydronic charge. It is not the potable vessel and never contacts drain water or potable water; it carries heat from one coil to the other.
Because the stored water is never on the potable side, the water age and temperature concerns that attach to potable storage do not attach to the vessel. Load shifting hot water without creating a water quality problem covers the potable-side questions that remain at the final heater.
The drain-side loop need not carry drain water. The defensible laundry design puts a trough or shell exchanger on the drain, as the Holiday Inn did, and runs a clean intermediate loop from that exchanger to the tank coil. Lint, grease and solids remain on the drain side of a serviceable exchanger rather than inside a corrugated coil.
Plumbing codes impose separation requirements wherever potable water is heated by a non-potable medium. The arrangement, wall count, heat transfer fluid and backflow provisions are decisions for the plumbing engineer and the authority having jurisdiction. This article makes no claim about which arrangement a given jurisdiction accepts.
The Btu balance for a stated laundry load
Every figure below is a stated input. Replace each with the facility’s own metered values before presenting any sizing or payback figure.
Case A assumes a laundry plant processing 1,200 lb per hour at 2 gallons per pound, the low end of EPA’s washer-extractor range. Blended drain water leaving at 120 F and cooled to 70 F across the exchanger are assumed temperatures. The calculation also assumes 50 percent exchanger effectiveness and an 8-hour shift.
V_drain = laundry throughput × water use per pound
V_drain = 1,200 lb/hr × 2 gal/lb
= 2,400 gal/hr
Q_avail = V_drain × 8.34 lb/gal × 1 Btu/lb-F × (T_drain − T_discharge)
Q_avail = 2,400 gal/hr × 8.34 lb/gal × 1 Btu/lb-F × (120 F − 70 F)
= 1,000,800 Btu/hr
= 293.3 kW thermal
Not all available heat transfers. The assumed exchanger effectiveness of 50 percent matches the ORNL triplex unit and sits at the low end of ORNL’s 45 to 85 percent laboratory range.
Q_rec = ε × Q_avail
Q_rec = 0.50 × 1,000,800 Btu/hr
= 500,400 Btu/hr
= 146.7 kW thermal
E_shift = Q_rec × t_shift
E_shift = 500,400 Btu/hr × 8 hr
= 4,003,200 Btu
= 1,173 kWh thermal per shift
In Case A, fill and drain are nearly in phase at the scale of the shift. The preheat tank buffers batch dumps rather than storing a full day’s recovery. The buffer calculation assumes one hour of recovery at a 40 F usable ΔT, charged to 100 F by the drain loop and depleted to 60 F by incoming makeup.
V_buffer = Q_rec × 1 hr / (8.34 lb/gal × 1 Btu/lb-F × ΔT_usable)
V_buffer = 500,400 Btu / (8.34 lb/gal × 1 Btu/lb-F × 40 F)
= 1,500 gal
At that 40 F usable ΔT, two 700-gallon modules hold 1,400 gallons, or 56 minutes of recovery; three hold 2,100 gallons, or 84 minutes. The longest gap between washer discharges the plant actually runs decides between them. Commercial Laundry Hot Water Systems sizes the hot side of the same plant.
The preheat tank is sized on the phase mismatch
Case B is a hotel with an on-premise laundry processing an assumed 2,400 lb per day over an assumed 8-hour shift, or 300 lb per hour. It uses the same assumed 2 gallons per pound, 120 F drain temperature, 70 F discharge temperature and 50 percent exchanger effectiveness. Guest showers are concentrated in an assumed 3-hour morning window that ends before the laundry starts. No heat recovered during the shift can be used at the moment it is recovered; all of it has to be stored overnight.
Q_rec = ε × (laundry throughput × water use per pound) × 8.34 lb/gal × 1 Btu/lb-F × (T_drain − T_discharge)
Q_rec = 0.50 × (300 lb/hr × 2 gal/lb) × 8.34 lb/gal × 1 Btu/lb-F × 50 F
= 0.50 × 600 gal/hr × 8.34 lb/gal × 1 Btu/lb-F × 50 F
= 125,100 Btu/hr
= 36.7 kW thermal
E_shift = Q_rec × t_shift
E_shift = 125,100 Btu/hr × 8 hr
= 1,000,800 Btu
= 293.3 kWh thermal per day
The preheat tank must hold the entire shift’s recovery at the assumed 40 F usable ΔT, charged to 100 F against a 120 F drain loop and depleted to 60 F against incoming makeup.
V_store = E_shift / (8.34 lb/gal × 1 Btu/lb-F × ΔT_usable)
V_store = 1,000,800 Btu / (8.34 lb/gal × 1 Btu/lb-F × 40 F)
= 3,000 gal
E_module = V_module × 8.34 lb/gal × 1 Btu/lb-F × ΔT_usable
E_module = 700 gal × 8.34 lb/gal × 1 Btu/lb-F × 40 F
= 233,520 Btu
= 68.4 kWh thermal at a 40 F usable ΔT
The module’s published nameplate is 108 kWh thermal at a 63 F ΔT, not at the 40 F usable ΔT used here. At 40 F, four modules give 2,800 gallons and 273.8 kWh thermal, 93 percent of the shift’s recovery. Five give 3,500 gallons and 342.2 kWh thermal, covering the full shift with margin for a longer laundry day.
Published standing loss on the 700-gallon module is about 2.4 F per 24 hours. The 14 hours between the end of the laundry shift and the start of the shower peak costs roughly 1.4 F, or about 3.5 percent of the 40 F usable band. The thermal storage tank sizing calculator runs the same arithmetic for other volumes and deltas.
During the 3-hour morning window, stored heat is released into the potable coil. On 3,000 gallons of morning makeup, the preheat carries the first 40 F of a 60 F lift from 60 F to 120 F. The final heater carries the remaining 20 F. Recovered heat displaces heater output only up to the preheat temperature.
P_displaced = E_shift / t_window
P_displaced = 293.3 kWh thermal / 3 hr
= 97.8 kW thermal
Thermal kWh is not electrical kWh. The following conversion assumes the final heater is a heat pump at a COP of 3.0. Replace that assumption with the rating at the design lift and the morning ambient.
E_elec = E_shift / COP
E_elec = 293.3 kWh thermal / 3.0
= 97.8 kWh electrical per day
P_elec = P_displaced / COP
P_elec = 97.8 kW thermal / 3.0
= 32.6 kW electrical during the morning window
If the final heater is a gas boiler at an assumed 80 percent efficiency, the same 1,000,800 Btu thermal displaces fuel input as follows.
E_fuel = E_shift / η_boiler
E_fuel = 1,000,800 Btu thermal / 0.80
= 1,251,000 Btu of fuel input
= 12.5 therms per day
E_elec = E_shift / COP
E_elec = 1,173 kWh thermal / 3.0
= 391 kWh electrical per shift
The ROI calculator walkthrough converts either figure to dollars against the facility’s own tariff. Thermal Energy Storage for Hotels covers the hotel plant beyond the laundry drain.
Equal flow vs heater only, and why storage removes the penalty
Efficiency Vermont’s installer guidance distinguishes two ways to plumb a gravity-film exchanger. In equal-flow plumbing, preheated water feeds both the water heater inlet and the cold side of the shower valve, so the full cold supply passes through the coil at the same rate the drain carries. In heater-only plumbing, preheated water feeds only the water heater inlet. The guidance states that heater-only plumbing reduces savings by about 20 percent.
CSA B55.1:20, the test method for efficiency and pressure loss of drain water heat recovery units, applies to units where potable and drain flow rates are equal. A rated efficiency therefore describes the equal-flow case and overstates a heater-only installation.
The penalty is a flow-balance penalty, not a plumbing-topology penalty. Heater-only plumbing starves the coil side because the heater’s cold makeup is a fraction of the drain flow. Applied to the restaurant pilot, a 26 percent equal-flow recovery falls to roughly 21 percent heater-only.
Applied to Case B, 293.3 kWh thermal per day would fall to roughly 235 kWh thermal, a loss of about 59 kWh thermal. At the assumed COP of 3.0, 59 kWh thermal / 3.0 is roughly 20 kWh electrical per day.
A stored preheat is heater-only by topology and equal-flow by physics. The drain side runs at the pumped rate the designer sets, the tank absorbs the heat, and the potable coil releases it at the draw rate. Recovery is limited by exchanger effectiveness and tank volume rather than by the ratio of two instantaneous flows.
Refrigeration heat reclaim for commercial hot water reaches the same conclusion from a different source: when source and load are out of phase, the tank makes recovery usable.
What the recovered heat is worth
Regulatory and market timing, as of the publish date
CSA B55.1:20, the third edition of the test method for drain water heat recovery unit efficiency and pressure loss, was reaffirmed in 2025 and remains the rating basis manufacturers cite as of October 2026. A rated efficiency under that method describes equal, simultaneous flow and should not be applied directly to a batch-load installation.
Model plumbing codes contain provisions governing heat exchangers that separate potable water from a non-potable heat transfer medium, and local amendments vary. The requirement that applies to a drain-side intermediate loop feeding a tank coil is the authority having jurisdiction’s determination, as of October 2026 and at any later date.
Grease interceptor requirements for restaurant drains are also set locally. Cooling drain water ahead of the interceptor is generally favorable to interceptor performance, but exchanger placement relative to a required interceptor is an AHJ question.
No incentive, tax credit or code eligibility claim is made in this article. Program rules change, and the reader’s utility and tax advisor make those determinations.
Checks before you specify a drain heat recovery system
Frequently Asked Questions
Does drain water heat recovery work for a commercial laundry?
Yes, with storage. DOE’s guidance states that a non-storage exchanger recovers heat only while drain and supply flow at the same time, and that clothes washers require storage capacity. A pumped drain-to-tank exchanger feeding a preheat tank decouples the two, as the ACEEE textile mill and Holiday Inn cases did.
How much heat can be recovered from laundry drain water?
At 1,200 lb per hour, 2 gallons per pound, a 120 F drain cooled to 70 F and 50 percent effectiveness, the recovered rate is 500,400 Btu per hour, or 146.7 kW thermal, and 1,173 kWh thermal per 8-hour shift. Every one of those inputs is an assumption to be replaced with metered data.
How big does the preheat tank need to be?
Large enough to hold the heat recovered during the hours when no hot water is drawn, at the ΔT the stored water can usefully swing. In the hotel case, 293.3 kWh thermal recovered over an assumed 8-hour laundry shift and released over an assumed 3-hour morning shower window requires 3,000 gallons at a 40 F usable ΔT. Four 700-gallon modules hold 93 percent of the shift’s recovery at that ΔT; five cover the full shift with margin.
What is the difference between equal-flow and heater-only drain water heat recovery?
Equal-flow plumbing feeds preheated water to both the heater inlet and the fixture’s cold side, so coil-side flow matches drain flow. Heater-only plumbing feeds the heater alone and starves the coil side; Efficiency Vermont puts the savings loss at about 20 percent. A stored preheat is not subject to that imbalance because the drain side is pumped at its own rate.
Is a wastewater-source heat pump better than a passive drain exchanger?
They do different jobs. A passive exchanger preheats to below drain temperature with no electrical input. A wastewater-source heat pump uses the drain as its evaporator source and lifts to delivery temperature, 140 F at an average COP of 3.7 in EPRI’s 60-unit pilot, at the cost of compressor input and equipment cost. Both used storage in the published commercial cases.
Does drain heat recovery raise a water quality concern?
The stored water in a coil tank is a closed hydronic charge that never contacts drain water or potable water; potable water passes through its own coil. The separation arrangement between potable water and the drain-side medium is governed by the plumbing code and decided by the authority having jurisdiction.
The tank is the component that makes commercial drain heat usable
Residential drain water heat recovery works without storage because the shower is its own load and its own source at the same instant. Commercial drains do not cooperate: laundry dumps in batches, the hotel’s drain runs in the afternoon while its showers run in the morning, and the dish machine drains in pulses. The heat is real and the published commercial paybacks are short, but the exchanger alone cannot deliver out-of-phase recovery to the load.
Size from the phase mismatch at the usable ΔT the drain temperature supports. Keep drain water out of the storage coil with a serviceable exchanger on the drain side, and rate the final heater for the lift that remains. Verify with the project’s engineering services and plumbing authority that the proposed solution will satisfy the project goals.
