A 2,379-square-foot house in San Francisco, standard plumbing bid, one-inch water service line sized per code, connection fee from the city: $26,000. Run the same fixture count through a probabilistic calculator built on data from 1,000 actual households, and the service line drops to three-quarter inch, the connection fee drops to $9,000, and the homeowner keeps seventeen thousand dollars that would have bought nothing but pipe capacity the house will never use.
A Formula Built for Fixtures That No Longer Exist
Roy B. Hunter was a physicist at the National Bureau of Standards, and in the 1940s he did something genuinely brilliant: he applied probability theory to plumbing, developing a statistical model that predicted peak water demand based on fixture counts and assumed simultaneous usage patterns. Engineers and plumbers adopted his curve into the Uniform Plumbing Code and the International Plumbing Code, where it has remained, with cosmetic adjustments, for more than eight decades.
Then Congress passed the Energy Policy Act of 1992, and toilets dropped from 3.5 gallons per flush to 1.6, faucets fell from 4 gallons per minute to 2.2, showerheads went from 5 gpm to 2.5, and WaterSense labeling pushed flows lower still. Fixture water consumption fell by more than half in a single legislative stroke, but nobody updated the pipe sizing formula that determines how much copper and PEX goes into your walls.
Flow monitoring studies conducted over the past two decades, compiled by PMM Magazine in a 2025 continuing-education review, show that Hunter's Curve overestimates peak water demand by factors ranging from five to twenty-seven times actual observed flow rates in residential settings. Systems designed to code, using the fixture-unit method Hunter devised for 1940s-era plumbing fixtures, see maximum real-world demands that barely reach 10 to 20 percent of their theoretical capacity. Eighty-plus percent of the pipe you paid for carries nothing but standing water, most of the time, in most of the rooms.
Standing Water Is Not Just Waste
Oversized pipes do not sit there quietly being expensive; they breed conditions that public health agencies have spent decades warning about. Water moves slowly through pipes that are too large for actual demand, and slow-moving water stagnates, ages, and loses its disinfectant residual, which is the chlorine that keeps pathogens from colonizing your plumbing. Biofilm forms on interior pipe walls, and biofilm is where Legionella pneumophila thrives, the bacterium responsible for Legionnaires' disease.
The CDC's guidance on controlling Legionella in building water systems specifically flags dead legs and oversized piping as risk factors. The EPA notes that stagnant or standing water in premise plumbing promotes Legionella growth, particularly after periods of low occupancy, and identifies pipe insulation and elimination of dead legs as design recommendations. One researcher at a 2020 EPA workshop put the dynamic plainly: once Legionella is in your house, it is the homeowner's responsibility, not the utility's.
There is a less dramatic but equally persistent consequence: you wait longer for hot water. Bigger pipes hold more cold water between the heater and the tap. Every morning, you run the faucet, watch money spiral down the drain, and wonder why a brand-new house takes forty-five seconds to deliver warm water to the kitchen sink. You paid for that delay, literally, in the oversized copper and PEX your plumber installed because an eighty-six-year-old formula told them to.
Two Tools That Actually Use Data
IAPMO's Water Demand Calculator is the first significant update to pipe sizing methodology since Hunter published his curve. Developed by the International Association of Plumbing and Mechanical Officials alongside the American Society of Plumbing Engineers and the University of Cincinnati, the WDC is built on field data from more than 1,000 single-family households. Researchers installed portable data loggers on each home's main water supply line and recorded every water-use event, every ten seconds, capturing 863,000 discrete events. They measured when people actually use water, how much, and for how long.
What they found demolishes the assumption of simultaneous fixture use. Faucets accounted for 75 percent of all water-use events but only 21 percent of total volume. Showers and clothes washers each represented fewer than 3 percent of events but roughly 24 percent of volume apiece. Real households do not behave the way Hunter's binomial distribution assumes. The WDC provides up to a 65 percent reduction in water demand estimates compared to fixture-unit methods, according to Builder Online, translating to one-to-three pipe-size reductions on the service line. It is available as a free app on Android and iPhone, or as an Excel download, and is now codified as IAPMO/UPC/Appendix M-2021.
The ICC took a different approach with DRIPS, a simulation platform developed in collaboration with the University of Miami that models water usage across an entire building over thousands of simulated days, incorporating regional data to adapt pipe sizing to geography, climate, and local usage patterns rather than relying on a single national assumption derived from 1940s-era fixture behavior. A companion initiative called SPADE standardizes how new usage data feeds into the model, keeping it current as fixtures get more efficient and occupancy patterns evolve.
What the Savings Actually Look Like
IAPMO commissioned Stantec Architecture to calculate cost savings across three building prototypes in three markets. For a typical 2,379-square-foot single-family home, applying the WDC saves $2,000 or more in high-cost labor markets and can reach $5,000 depending on location. For a 45-unit multifamily building, savings exceed $100,000, particularly where utility meter connection fees reflect water scarcity pricing.
A separate study by the Alliance for Water Efficiency examined meter connection fees in the 50 fastest-growing U.S. metro areas and found that over 60 percent charge more than $1,500 for a one-inch meter connection, while over 40 percent charge more than $3,000. Dropping from a one-inch to a three-quarter-inch meter, which is exactly what the WDC enables for most single-family homes, eliminates that fee differential entirely.
Why Plumbers Haven't Switched
Fear, mostly. An undersized pipe during a dinner party is a catastrophe that every plumber can visualize and no plumber wants to explain, while an oversized pipe is invisible, costs the homeowner money they will never know they lost, and produces a health risk that takes years to manifest. The incentive structure favors caution, and caution here means big pipes.
There is also institutional gravity that should not be underestimated. Hunter's Curve has been embedded in plumbing codes for eight decades, and plumbing engineers learned it in school, applied it on every project, and passed it to apprentices who did the same without questioning whether the underlying assumptions still held. PMM Magazine's 2025 review noted that the industry "continues to prioritize perceived robustness over demonstrated performance" and that many engineers stick with legacy spreadsheets out of habit rather than evidence, because adopting the WDC means admitting that thousands of homes they already piped were overbuilt.
The Counterargument Worth Taking Seriously
Plumbers are not wrong to worry about edge cases. The WDC's dataset of 1,000 households captures typical single-family usage, but residential construction is not always typical. A four-bathroom house with a pool, a separate guest suite, and an irrigation system fed from the domestic supply presents a peak-demand profile that probabilistic averages may underestimate. Homes in regions with hard water and high mineral content experience flow restriction from scale buildup over time, which means a pipe that seems adequate at installation may be functionally undersized a decade later. And jurisdictions that have not yet adopted Appendix M leave builders in a bind: the WDC says three-quarter inch, the local code says one inch, and the inspector cares about the code, not the calculator.
What Builders and Homeowners Should Do
If you are building or bidding a residential project, download the IAPMO Water Demand Calculator and run your fixture count through it before your plumber hands you a materials list, then compare the WDC output to your plumber's sizing and ask which method they used if you find a gap of one or more pipe sizes.
If you are a homeowner wondering why your hot water takes forever, ask your plumber whether a recirculation loop or a point-of-use heater could solve the delivery delay that oversized trunk lines created, since both are cheaper than repiping and address the symptom directly.
If you are a jurisdiction, adopt Appendix M, because builders in your market are paying for pipe they do not need, homeowners are waiting for hot water they already paid to heat, and standing water in oversized systems is creating conditions the CDC has explicitly flagged as Legionella risk factors. The data exists, the tools are free, and the only thing missing is the political will to update an eighty-six-year-old default that has cost American homebuilders billions of dollars in unnecessary copper, PEX, and meter connection fees over the past three decades.
Limitations
The Stantec cost figures use 2021 construction pricing and three specific metro areas; actual savings vary by region, material choice (copper vs. PEX), and local labor rates. The 5-to-27x overestimate range aggregates multiple studies with different methodologies, building types, and climate zones; individual results depend on household size, fixture mix, and occupancy pattern. IAPMO's WDC is currently validated only for residential buildings; commercial applications are pending further research. DRIPS is operational but adoption data, including the number of jurisdictions or projects using it, is not publicly available. Legionella risk from oversized residential plumbing is established in the epidemiological literature for institutional and commercial buildings; direct single-family residential outbreak data attributable to pipe oversizing specifically is limited.