Water, Drainage, Gas, and Storm Sizing
The master plumber exam is built on calculation, and this is the chapter where most of the arithmetic lives. You are expected to move from a fixture-unit count to a pipe size, hold the right slope over a run, balance a pressure budget from the meter to the most remote fixture, convert an appliance BTU rating into gas flow and then into pipe size, and size roof drainage from a rainfall rate. Each of these follows a repeatable procedure: total the load, convert the load to a flow, and read the flow against a sizing table at the governing length or slope. Master the four conversions in this chapter, WSFU to gpm, DFU to drain size, BTU to cfh, and roof area to storm gpm, and most of the sizing questions become mechanical.
Water Supply Sizing by Fixture Units
Water distribution is sized from water supply fixture units (WSFU), a probability-weighted value the code assigns to each fixture so that dissimilar fixtures can be added on one scale. You total the WSFU carried by a section of pipe, then convert that single total to a design flow in gallons per minute using a demand curve derived from Hunter's probability work, which accounts for the low likelihood that every fixture runs at once. The single most tested rule is that you add fixture units first and convert the combined total to gpm only once; converting each branch to gpm and summing the flows overstates demand badly because the curve already discounts simultaneous use. Flush-valve (flushometer) fixtures and flush-tank fixtures ride different curves, because a flushometer draws a heavy momentary flow while a tank refills slowly through a small ballcock. The two model codes share this method but house the tables differently: the UPC prints demand in Table 610.3, while the IPC places equivalent fixture-unit values and the Hunter demand curve in Appendix E, which many jurisdictions adopt. Once you have the design gpm, the pipe diameter is chosen so velocity stays within limits, generally about 8 ft/s for cold water and 5 ft/s for continuously circulated hot water, computed from v = 0.408 x Q / d squared, where Q is gpm and d is the actual inside diameter in inches. High velocity drives erosion-corrosion, noise, and water hammer, and hot water is chemically more aggressive, which is why its ceiling is lower. A distribution main is also never reduced in the direction of flow toward a fixture group it still must feed: the main grows as it collects load, and a downstream reduction would starve every fixture past the necking-down. Minimum fixture supply pressures, often 8 psi for a lavatory and up to about 25 psi for a flushometer, feed forward into the pressure budget of the next section.
The Pressure Budget
Every water system balances the pressure available at the source against everything that consumes it before water reaches the most remote and highest fixture. The consumers are four: static lift, at 0.433 psi per foot of elevation gain; fixed device losses such as the water meter, softener, filter, and backflow assembly, each read from the manufacturer's chart at design flow; pipe friction, expressed as psi per 100 ft; and the minimum flow pressure the critical fixture itself demands, commonly 8 to 25 psi. The procedure is subtraction: start from the minimum guaranteed street pressure, subtract elevation, subtract meter and device losses, and subtract the fixture's required residual, and whatever remains is the pressure budget available to overcome friction. Divide that remaining pressure by the developed length of the critical run and multiply by 100 to get the allowable uniform friction loss per 100 ft. That per-100-ft value, carried across a friction chart at the section's design gpm, selects each pipe size so the whole run just spends its budget without exceeding it. If the friction loss at a trial size exceeds the allowable rate, the pipe is enlarged one size and rechecked, or a booster pump or pressure-reducing strategy is reconsidered where street pressure is too low or too high. A subtle but heavily tested point is developed length versus measured length: you must add the equivalent length of every fitting, valve, and the meter to the straight-pipe measurement, because a system rich in elbows and a globe rather than a full-port valve behaves as if it were far longer than a tape says. Worked backward, the same budget tells you the tallest building a given street pressure can serve by gravity before a booster is required, since roughly every 2.31 ft of rise costs 1 psi. Both the UPC and IPC accept this segmented friction-loss method; the IPC also permits a simplified prescriptive table for small systems.
Drainage and Vent Sizing by Fixture Units
Sanitary drainage is sized from drainage fixture units (DFU), a value that scales roughly with a fixture's discharge rate and its potential to load the drain. The DFU tables are split into columns that must never be interchanged: the horizontal fixture branch, the vertical stack (often qualified by the number of branch intervals), and the building drain and building sewer at a stated slope. You total the DFU a section carries and select the smallest pipe whose column capacity meets or exceeds that load. Two override rules bind regardless of the count: a drain may never be smaller than the largest trap or trap arm it serves, and any drain receiving a water closet must be at least 3 inches. Slope is the second axis of drainage sizing. Both model codes set a minimum of 1/4 inch per foot for pipe 2-1/2 inches and smaller and 1/8 inch per foot for 3 inches and larger, with 1/16 inch per foot allowed for very large sewers only by special approval; total fall is simply slope times developed run, so 96 ft of 4-inch drain at 1/8 inch per foot drops 12 inches. Too little slope lets solids strand; too much can let liquid outrun solids, though modern research has softened the old scouring-velocity fear. Venting is sized in parallel: a vent is not less than half the diameter of the drain it serves and never smaller than 1-1/4 inches, and its size is further limited by developed length so friction does not defeat it. IPC and UPC diverge on details worth memorizing: they assign slightly different DFU values to some fixtures, and the maximum trap-arm length before the vent (the distance a trap can sit from its vent) is tabulated by pipe size, roughly 6 ft for a 2-inch arm. Always confirm which code, and which edition, the jurisdiction has adopted.
Gas Pipe Sizing
Fuel-gas piping is sized by first converting each appliance input, rated in BTU per hour, into a volume flow in cubic feet per hour (cfh) by dividing by the fuel's heating value: about 1,000 BTU per cubic foot for natural gas and about 2,500 BTU per cubic foot for undiluted propane. Because propane packs roughly 2.5 times the energy per cubic foot, the same BTU load produces far fewer cfh and often a smaller pipe, though propane's lower delivery pressure and its own tables must be used. The whole-house connected load is the sum of every appliance input, and it sizes the meter and the section of main upstream of the first tee. Individual runs are then sized by the longest-length method: measure the single longest run from the point of delivery to the most remote outlet, and size every section of the system from the table column for that one length, which conservatively covers worst-case simultaneous demand. A branch upstream of any tee always carries the sum of all downstream appliance demands, while each appliance connector is sized only for its own load. The governing tables, in NFPA 54 (the National Fuel Gas Code) and the adopted IFGC, are indexed by pipe material (schedule 40 steel, copper, CSST, or polyethylene), by operating pressure and the allowable pressure drop, and by specific gravity. A worked example: a 120,000 BTU/hr furnace on natural gas is 120 cfh; add a 40,000 BTU/hr water heater (40 cfh) and a 65,000 BTU/hr range (65 cfh) and the common section carries 225 cfh, which is then read at the longest length to pick the diameter. An alternative branch-length or hybrid method sizes each section by its own length and generally yields smaller pipe, but it is permitted only where the code and table specifically allow it.
Storm Drainage Sizing
Roof and storm drainage is sized from the locality's design rainfall rate, expressed in inches per hour for a stated return period (commonly the 100-year, one-hour storm), applied to the projected horizontal roof area that drains to a given conductor. The rational method, Q = 0.0104 x A x i, converts area A in square feet and intensity i in inches per hour into a design flow in gpm, where the constant 0.0104 is the flow produced by one inch per hour of rain falling on one square foot. That flow is read against leader (vertical conductor), horizontal storm-drain, roof-drain, and gutter tables to size each component. Because published tables are usually printed at a reference rate such as 4 inches per hour, the allowable roof area a given conductor can serve scales inversely with the local rate: halve the rainfall intensity and a conductor drains twice the area; double the intensity and it drains half. A worked example: a 6,000 sq ft roof at 3 inches per hour produces 0.0104 x 6,000 x 3 = 187 gpm, which sets the leader size from the table. Vertical walls that shed onto a roof add 50 percent of the wall area (for one adjacent wall) to the contributing area in many tables. Secondary, or overflow, drainage is mandatory: an independent overflow drain or scupper set above the primary drain's inlet keeps a blocked primary from ponding water, since standing water weighs about 5.2 psf per inch of depth and can overload or collapse the structure. The UPC houses storm sizing in Chapter 11 and the IPC in its own Chapter 11, but the specific tables, the sizing of controlled-flow roof drains, and the overflow provisions differ in detail, so size to the adopted edition.
Last updated: September 2026
