Water supply pipe sizing and WSFU field guide for plumbers

Water supply sizing and what it has to deliver

Water supply pipe sizing is how you set the diameter of every pipe in the potable distribution system, from the service at the meter to the last stub at the fixture, so that each outlet gets enough flow at enough pressure when the building is busy. It is the water side of the building's plumbing. The drain side, sized by drainage fixture units, is the companion problem and runs on its own rules, covered in the DWV and venting guide.

Two numbers fight you the whole way: flow and pressure. You need enough flow, in gallons per minute, to feed every fixture likely to be open at once, and you need enough pressure left at the worst fixture to actually push that flow through it. Pipe that is too small starves the far end. Every foot of pipe and every fitting eats pressure as friction, and the smaller the pipe at a given flow, the faster the water moves and the more pressure it burns.

So sizing is a budget problem, not a single formula. You start with the pressure the source gives you, you subtract everything that takes a cut before the water reaches the fixture, and you size the pipe so the friction left to spend gets you to the most remote outlet with the residual pressure that fixture needs still on the gauge. Get any input wrong, the routed length, the demand, the elevation, and the answer looks fine on paper and runs weak in the building.

What too small and too big each cost you

Undersized water pipe announces itself the day two fixtures run together. The shower goes weak when someone flushes, the hose bibb drops the upstairs lav to a trickle, and the dishwasher times out filling because the supply could not keep up. Each fixture passed its own test alone. Sized for the real peak, they fail together, because the pipe could not carry the simultaneous demand without the pressure collapsing on the longest run.

Oversized pipe fails more quietly and people miss it. It costs more up front in material and labor, and that is the part estimators see. The part they do not see is water age. Water sitting in oversized pipe moves slowly and turns over rarely, the disinfectant residual decays, and the dead volume becomes a place for biofilm and Legionella to set up, especially on the hot side and on low-use branches. Bigger is not safer on potable water. There is a window, and you are aiming for the middle of it.

The target is the smallest pipe that carries the peak demand with the residual pressure intact at the far fixture and the velocity under the erosion limit. That is the whole job. Hit it and the system delivers and lasts. Miss it small and you starve fixtures. Miss it big and you waste money and grow water-quality problems you will not see until someone gets sick or the copper pinholes.

What is a water supply fixture unit?

A water supply fixture unit, the WSFU, is a dimensionless number assigned to a fixture that stands for its load on the water supply, folding together how much it draws, how long it draws, and how often it gets used. It is a load currency, not a flow rate you measure with a gauge. A fixture that dumps a lot of water fast and gets used often carries more WSFU than one that sips.

The unit traces back to Roy Hunter's work at the National Bureau of Standards around 1940, which turned a building full of unlike fixtures into a common load value so a designer could add them up and estimate a realistic peak. The genius of it is that the unit already carries the probability of use baked in, so totaling WSFU and converting to gpm gives you a probable simultaneous demand, not the impossible sum of every fixture wide open.

Code tables assign each fixture type a WSFU value, and the values differ between private use, a fixture serving one dwelling or one family, and public use, a fixture in a building open to many people, which is hit harder and carries more units. Verify the values against the adopted code, because the IPC and UPC table sets are not identical, but the common ones are stable. The table lists the everyday values a plumber and an estimator carry in their head.

Why not size for every fixture running at once?

You do not size for every fixture at full flow because that moment never happens, and designing for it buries the building in oversized pipe that then sits half-stagnant. A school does not flush every closet and open every lav in the same second. The odds that all of them coincide are vanishingly small, and they fall as the building gets bigger. That statistical reality is called diversity, and it is the single fact that makes water sizing economical instead of absurd.

Hunter's method captures the diversity in the curve that converts fixture units to demand. Add a fixture and the total WSFU climbs in a straight line, but the probable gpm demand climbs along a curve that flattens hard as the count grows. Ten fixtures might run at a real demand close to their summed flow, because with so few, a coincidence is plausible. A thousand fixtures run at a demand far below their summed flow, because the odds of them all opening together have collapsed. The curve, not the sum, is the demand.

This is also why you never just add up the rated gpm on every fixture and call it the load. That number is real for one fixture and meaningless for a building. The WSFU total run through the demand curve is the probable peak the system actually has to serve, and it is the number the pipe gets sized to.

From fixture units to gallons per minute

Once you have the total WSFU for a segment, you convert it to a probable demand in gpm using the Hunter demand curve or the equivalent table in the adopted code, commonly the table in IPC Appendix E. You read the total fixture units in, you read the probable demand out. The relationship is not linear, so doubling the fixture units does not double the gpm. The curve does the diversity math for you.

The curve splits into two lines at the low end, and which one you read matters. The upper line is for systems with flush valves, the lower line for systems with flush tanks. The split exists because the two flush differently. A flushometer valve dumps roughly 4 gallons in about 9 seconds, a design flow near 27 gpm in a hard, short burst. A flush tank meters the same 4 gallons in around a minute, a design flow closer to 4 gpm, because the tank refills slowly through a small valve. A building full of flushometers has a spikier, higher short-duration demand than the same building on tanks, so it reads off the higher line and often needs larger pipe and a hard look at the supply pressure.

Use the demand value the adopted code's table gives for your fixture-unit total and your flush type. The exact numbers in the curve have been revised over the years as fixtures got more efficient and the old Hunter values started to overstate modern low-flow demand, so the table in the code edition you are sizing to is the one that governs, not a curve copied from an old handbook.

The pressure budget and the friction left to spend

Sizing runs off a pressure budget. You start with the pressure the source gives you, the minimum daily service pressure the water authority guarantees at the meter, not the peak it sometimes reads. Then you subtract every fixed loss between the source and the worst fixture, and what survives is the pressure you are allowed to spend on pipe friction. Size the pipe so the friction over the longest run does not exceed that survivor, and the system works.

Four things take their cut before friction. Elevation: lifting water up a building costs about 0.43 psi for every foot of height, the static head, and a precise figure is 0.433 psi per foot, so a fixture 40 ft up has already lost about 17 psi before any flow starts. The water meter takes a loss that grows with the flow through it, read from the meter manufacturer's curve. A backflow preventer, where the service has one, takes a real and often large loss, commonly several psi to over 10 psi on a reduced-pressure assembly, read from the device data. And the worst fixture itself demands a minimum residual flowing pressure to operate, which you must leave on the table.

Subtract all four from the source pressure and the remainder is the pressure available for friction, the number that sizes the pipe. Spread that remainder over the developed length of the longest run, including the equivalent length of the fittings, and you get a permissible friction loss per 100 ft. Size every segment on that run to stay at or under that figure and the far fixture gets its flow at its residual pressure. A common planning move is to hold friction to no more than about 5 psi per 100 ft, but the real ceiling is whatever the budget leaves, which on a tall building or a long site run can be a good deal tighter.

P_supply

Minimum daily service pressure at the source, in psi, the guaranteed low, not the peak

P_elev

Static head loss lifting water to the highest fixture, about 0.43 psi per foot of rise

P_meter / P_bf

Pressure loss through the water meter and any backflow preventer at the design flow, from the device data

P_res

Minimum residual flowing pressure the worst fixture needs to operate, left on the budget

L_dev

Developed length of the longest run in feet, including the equivalent length of fittings

What velocity is too high for water pipe?

Velocity is the second ceiling, and it caps the pipe independently of the pressure budget. The common design limits are about 8 ft per second for cold water and about 5 ft per second for hot, with the hot limit lower because hot water is more aggressive and erodes the pipe wall faster. Push past these and you get erosion corrosion, where the moving water strips the protective film off the inside of the pipe and grooves out the elbows and tees, plus noise and the conditions for water hammer.

The limits tighten with the water and the temperature. Copper carrying water above 140 degrees F should drop to roughly 2 to 3 ft per second, because the hotter the water the more aggressively it attacks the wall. Aggressive, low-pH water, below about 6.9, calls for holding even cold velocity down near 4 ft per second. These are commonly cited engineering limits rather than a single hard code number, and the adopted code and the copper tube handbook give the figures, so verify them for the material and the water chemistry on the job.

The failure mode is specific and you can read it. Erosion corrosion shows up as pinholes and thinned, scalloped grooves on the inside of elbows and on the downstream side of fittings, always where the flow turns and the velocity spikes. A hot recirculation loop run too fast is the classic place copper erodes through years early. That same high velocity is what slams the pipe when a valve shuts fast, the water-hammer angle, so holding velocity down buys you quiet pipe and long-lived fittings at the same time.

Friction loss and sizing each segment

With the permissible friction per 100 ft in hand, you size the run one segment at a time. A segment is a stretch of pipe carrying a constant flow, between the points where fixtures or branches join and the demand changes. You work from the source toward the far fixture, and at each segment you know the WSFU downstream of that point, which gives you the demand in gpm for that segment off the demand curve.

For each segment you have the flow and you have the friction budget per 100 ft, and you read the pipe size off a friction chart or table for the material that delivers that flow at or under that friction loss while keeping the velocity under the limit. Two constraints, one size. The chart by material and flow gives friction per 100 ft and velocity together for each diameter, so you pick the smallest size that satisfies both. The velocity limit often governs on the larger flows, the friction limit on the longer runs.

Developed length is what people get wrong, the same mistake as on the drain side. It is the measured length the pipe actually runs, up and over and around, plus the equivalent length the fittings add, not the straight-line distance on the plan. Every elbow, tee, and valve behaves like extra feet of pipe, and on a fitting-heavy run those equivalent feet add up to a real fraction of the loss. Measure the route, add the fittings, then check that the total friction down the longest run fits inside the budget. If it does not, you go up a size on the segments that are burning the most, usually the long mains and the busiest branches.

Copper, PEX, and CPVC and how they size differently

The material changes the size, because the same nominal pipe has a different inside diameter and a different friction character in copper, PEX, and CPVC. Smooth bore and a bigger inside diameter mean less friction at a given flow, which can let you hold a size. A rougher or narrower bore burns more pressure and can push you up a size to keep the far fixture fed.

New copper, PEX, and CPVC all start with a similar friction coefficient, a Hazen-Williams C around 150, so on the straight pipe alone they are close when new. Two things separate them. Copper's C falls over time as the inside scales and corrodes, so an old copper system runs rougher than the chart says, while CPVC holds its smoothness. And PEX in the common copper-tube-size dimension has a thicker wall and therefore a smaller inside diameter than copper of the same nominal size, so PEX runs a higher velocity and more friction at the same flow and often wants a size up on a long or busy run.

Fittings are the bigger PEX story. Insert-style PEX fittings neck the bore down at every joint, and the flow restriction through them is real, far more than through a copper or CPVC fitting, so the equivalent length you add for PEX fittings is larger and a fitting-heavy PEX run loses more than the straight-pipe number suggests. The fix is to use the larger-bore PEX fitting systems where the run is demanding, or to go up a pipe size, and to size PEX off the manufacturer's own flow and pressure-drop data rather than a copper chart. The velocity limits still apply to all three, and on the hot side the 5 ft per second ceiling holds regardless of material.

The service, the meter, the mains, branches, and risers

A distribution system is a chain of segments, each with its own job and its own minimum. The water service is the buried pipe from the main to the building, sized for the whole building's peak demand plus any losses out at the street, and coordinated with the meter and the water authority's tap rules. The meter sits in that line, and its size sets a flow ceiling and a pressure loss of its own, so an undersized meter chokes the building no matter how generous the pipe behind it.

Inside, the water-service pipe becomes the building main, the horizontal artery feeding the risers. A riser is a vertical run carrying water up through the floors, and a branch is the horizontal run off a riser or main feeding a group of fixtures. Each one is sized for the WSFU downstream of it, so the main carries the most and the last fixture branch the least. The demand shrinks as you move out toward the fixtures, and the pipe steps down with it.

Minimum sizes set a floor under the calculation, and they are not optional. A fixture supply, the last stub to the fixture, has a code minimum by fixture type, commonly 1/2 in for a lavatory or sink and 3/8 in for some small fixtures, while a flushometer water closet needs a 1 in supply because of its hard, short burst. The calculation can call for bigger. It can never go below the minimum, and an inspector checks the fixture supply sizes against the code table as a matter of routine. Verify the minimums against the adopted code, because the IPC and UPC lists differ in places.

How much flow and pressure does each fixture need?

Every fixture has two numbers that drive the far end of the budget: the flow it needs in gpm, and the residual flowing pressure it needs to deliver that flow. The residual is the pressure that has to still be on the gauge at the fixture with the water running, and it is what you reserve before you spend anything on friction. Size the run so the worst fixture keeps its residual, and the easier fixtures upstream are fine by default.

Flush tanks are forgiving. A flush-tank water closet commonly needs only about 8 psi flowing residual, because it fills slowly into a tank and does not care much about pressure. Flushometer valves are the hard case. A flushometer needs a much higher residual to snap through its cycle and deliver the flush, commonly in the range of 15 to 35 psi flowing depending on whether it is a siphonic or a blowout closet and on the valve, with the IPC residual table, commonly Table 604.3, giving the figure. That high residual is why a flushometer building can be a pressure problem on the upper floors even when a flush-tank building on the same service would be fine.

Carry the fixture residuals as the hard floor of the budget. A private lavatory commonly needs about 0.8 gpm at 8 psi, a shower runs at its rated flow at a modest residual, and the flushometer closet sets the high bar. The maximum side matters too: code caps the static pressure at the fixture, commonly 80 psi, above which you must put in a pressure-reducing valve, because high pressure wears fittings, worsens water hammer, and makes thermal expansion worse on the hot side. Verify the residuals and the maximum against the adopted code, since the table is what an inspector reads.

Field example: a small commercial run, segment by segment

Take a two-story commercial building on a 60 psi minimum daily service pressure, with the highest fixture about 25 ft above the meter and a 3/4 in displacement meter in the line. Walk the budget first, then size the pipe. The elevation eats 0.433 times 25, about 10.8 psi. The meter at the building's demand reads, say, 6 psi off its curve. There is no backflow assembly on this service. The worst fixture is a flush-tank closet needing 8 psi residual.

Start with 60 psi, subtract 10.8 for lift, 6 for the meter, and 8 for the residual, and you are left with about 35 psi to spend on friction across the whole run to that closet. Measure the developed length to the worst fixture, say 180 ft of pipe, then add the equivalent length of its fittings, call it another 70 ft, for 250 ft developed. The permissible friction is 35 psi over 250 ft, which is 14 psi per 100 ft. That is generous, so velocity will likely govern the size here, not friction.

Now size each segment to its own demand. Total the WSFU downstream of each point, read the gpm off the demand curve, and pick the smallest pipe that carries it under both the 14 psi per 100 ft friction figure and the 8 ft per second cold velocity limit. The main carries the building total and lands largest. The branches and the fixture supplies step down to the minimums as the demand falls toward the last lav. Change one input, drop the service to 45 psi or add a reduced-pressure backflow assembly, and that friction budget tightens fast, the pipe goes up a size on the long segments, and a borderline run can need a booster.

The hot side and the recirculation tie

The hot-water distribution sizes the same way as the cold, off WSFU and the pressure budget, with two differences that change the answer. The velocity limit drops to about 5 ft per second because hot water erodes copper faster, so the hot side often wants a slightly larger pipe than the cold for the same flow. And the hot demand is a fraction of the total, because not every fixture draws hot, so the hot WSFU on a segment is lower than its combined count.

On any run long enough that the wait for hot water gets unreasonable, the design adds a recirculation loop, and the loop changes the hot-side sizing. The supply still sizes for fixture demand, but now there is a return pipe carrying the cooled loop water back to the heater, sized for the small recirculation flow that makes up the loop's heat loss, not for any fixture draw. That return flow is only a few gpm on most loops, so the return pipe is small, but it has to be balanced across the risers or the far branch runs cold. The recirculation loop, the pump sizing, and the temperature controls are their own design, covered in the water heater recirculation guide.

The trap to avoid is running the recirculation loop velocity too high to force the temperature out to the far riser. Copper carrying hot water above the velocity limit erodes through years early, right at the elbows, and the loop is where that happens because it runs all day. Hold the loop velocity down, insulate the pipe to cut the heat loss, and balance the flow, rather than cranking the pump and eating the pipe.

High-rise: pressure zones, PRVs, and booster pumps

On a tall building the elevation head wins the budget by itself. At 0.43 psi per foot, a fixture 200 ft up has lost about 86 psi to lift alone, more than most street pressure can give, so the street cannot serve the top and a booster system has to make up the difference. The design splits the building into pressure zones up its height, each zone served at a pressure that keeps its top fixtures fed and its bottom fixtures under the maximum.

The reason you zone instead of just boosting the whole stack to the top is the maximum-pressure cap at the fixture, commonly 80 psi. Boost the base of the building hard enough to push water to the top floor and the bottom floors would see far over 80 psi, which wears fittings and slams the pipe. So the building is broken into zones of a workable height, each zone gets the pressure it needs at its top, and a pressure-reducing valve knocks the pressure back down within a zone where it would otherwise run too high at the lower floors. The PRV is what keeps the bottom of a zone under the cap.

The boost itself comes from a booster pump system, sized for the building's peak demand at the head needed to reach the top zone, commonly a set of pumps with variable speed so the system tracks the real demand instead of running flat out. Some tall buildings still use a gravity tank up top filled by pumps, drawing down through the day. Either way, the booster, the zones, and the PRVs are an engineered system on a high-rise, designed and stamped, with the demand still coming from the WSFU total run through the curve. The fixture-unit method sizes the pipe inside each zone the same as any building.

What method does the code use to size water pipe?

The adopted plumbing code carries the sizing method, and under the IPC it lives in Appendix E, which the jurisdiction must specifically adopt because appendices are not automatically in force. Appendix E lays out more than one procedure. The thorough one is the segmented loss method, which is the pressure-budget approach written out: it sizes each segment from the water supply demand in gpm, the available pressure, and the friction loss from the meter and the developed length including the equivalent length of fittings, summing the losses so the total does not exceed the pressure available at the source.

There is also a simpler prescriptive path in the code for ordinary buildings, a table that gives a pipe size directly from the fixture-unit count and the available pressure range and developed length, so a small job does not need the full segment-by-segment calculation. Use the prescriptive table where the building fits its limits, and run the segmented method where the building is large, tall, or tight on pressure. The UPC handles water sizing through its own water-distribution provisions and tables, which differ from the IPC, so the method as well as the numbers depends on which code your jurisdiction adopted.

Do not cite a specific section or table number from memory across editions. The appendix designation, the table numbers, and the demand figures shift between code cycles, and Appendix E in particular is only in force where it has been adopted. Confirm the method, the tables, and the adoption against the edition the jurisdiction enforces and any local amendments before you size off it on a permit set.

Oversizing, stagnation, and water age

There is a hard floor on how big you should go, and it is a water-quality floor, not a cost one. Oversized pipe holds water that turns over slowly, and slow turnover means high water age, the time water sits in the system between entering and being used. As water ages the disinfectant residual decays, the temperature drifts toward the range bacteria like, and the dead volume on a low-use branch becomes a reservoir for biofilm and Legionella. The same instinct that makes a plumber round up a size to be safe is the instinct that grows the problem.

It bites hardest in a few places. The hot side and the recirculation loop, because warm water is in the growth range to begin with. Oversized branches to fixtures that rarely run, like a janitor's sink or a fixture in a space that sits empty. And buildings that were sized for a future occupancy that never filled in, so the pipe is carrying a fraction of the demand it was sized for and the water barely moves. Sizing tight, to the real demand, keeps the water moving and the residual fresh.

Where a branch has to serve a low-use or seasonal fixture, the answer is not oversized pipe, it is right-sized pipe plus a flushing plan to turn the water over, and on a higher-risk building a water management program that monitors and manages it. The plumbing side of that program starts with not building stagnant dead volume in the first place. Size to the demand, keep dead legs short, and do not round up out of habit on a low-use run.

Large commercial and data center domestic water

On large commercial work the WSFU method still sizes the pipe, but the demand mix shifts and a few loads do not fit the fixture-unit table at all. A building with heavy continuous or process demand, a commercial kitchen, a laundry, a cooling-tower makeup line, or an irrigation tap, has loads that run steadily rather than intermittently, and those do not diversify the way fixtures do. You add the continuous demand to the diversified fixture demand rather than running it through the curve, because the curve assumes intermittent use and a steady load breaks that assumption.

A data center is the sharp version of this. The fixture count for the people is small, but the domestic and makeup water for humidification and for evaporative or adiabatic cooling can dwarf the restroom demand, and that load is closer to continuous. Sizing the service off the restroom fixture units alone would badly undersize it. The mechanical makeup demand is its own number, often the governing one, and the domestic potable system is designed around both the small diversified fixture load and the large steady mechanical load together, frequently with separate metering and backflow protection on the makeup line so it cannot contaminate the potable supply.

Across all the large work, two things stay true. The diversified fixture demand still comes from the WSFU total and the demand curve, and the continuous loads get added on top from their own rated flows. And the bigger and more critical the building, the more the design is an engineered, stamped system run off the project documents, with the fixture-unit method as the framework for the fixture portion, not the whole answer.

What to document

If the segment math is not on file, nothing shows the pipe was sized to the demand against the real pressure once the far fixture runs weak or the inspector asks where the service size came from. The record is what proves the pipe was sized to the demand against the real pressure, and it is what the next plumber reads when the building gets a tenant fit-out that adds fixtures to a branch that was sized tight.

Build the record as a segment table. For each segment capture the WSFU total downstream, the demand in gpm off the curve, the developed length including fittings, the pipe size, the velocity at that size, and the running pressure loss, alongside the pressure budget that set the whole thing: the source pressure, the elevation, the meter and backflow losses, and the worst-fixture residual. Record which code, edition, and method you sized to, because the demand numbers only mean something tied to the table that produced them.

Common mistakes

• Adding up every fixture's full rated flow with no diversity, instead of running the WSFU total through the demand curve.
• Ignoring the elevation head and the meter and backflow losses, so the budget looks bigger than it is and the far fixture runs weak.
• Sizing pipe small enough to push velocity past the limit, eroding copper at the fittings and inviting water hammer.
• Undersizing a branch or a flushometer supply below the demand and the code minimum, so it starves when a second fixture opens.
• Reading the flush-tank demand line for a flushometer building, understating the spiky short-duration peak.
• Rounding every pipe up a size out of habit, building stagnant dead volume that ages the water and grows Legionella.
• Using the plan distance instead of the developed length plus the fitting equivalent length.
• Sizing off an old Hunter curve or the wrong model code instead of the demand table in the adopted edition.

Field checklist

Standards and references

The framework lives in the two model plumbing codes, and the adopted edition with local amendments controls. The IPC, published by the ICC, carries water supply and distribution in its Chapter 6 and the sizing procedures in Appendix E, which a jurisdiction must specifically adopt for it to be in force. The UPC, published by IAPMO, covers water sizing through its own water-distribution provisions and tables. The fixture-unit values, the demand conversion, the minimum residual pressures, and the minimum pipe sizes all sit in those code chapters, and the section and table numbers shift between editions, so confirm them against the edition the jurisdiction enforces before citing one on a submittal.

The method underneath the tables traces to the original fixture-unit and demand work by Roy Hunter at the National Bureau of Standards around 1940, which the model codes adapted and have revised over the decades as fixtures got more efficient and the early demand values began to overstate modern low-flow loads. ASPE design references give the engineering background behind the tables for large and non-standard systems, and AWWA covers the water service and the meter at the street. The velocity limits and the material behavior come from the copper tube handbook and the plastic-pipe manufacturers' design data for PEX and CPVC, which is where the per-material friction and fitting losses live.

The local water authority is the other governing party, because it sets the guaranteed service pressure you build the budget on, the meter and tap rules, and the backflow protection it requires at the service. Cite the standard that controls the point, confirm it against the adopted code and the water authority, and let the project documents and the equipment data override any rule of thumb in this guide.

Units, terms, and conversions

Water supply sizing carries a few units and a few names for the same idea, and the same value can read differently across a plan, a code table, and a manufacturer's chart.

Fixture load is counted in water supply fixture units, the WSFU, sometimes just called fixture units or FU on a drawing. Flow is in gallons per minute, gpm, and metric drawings use liters per second. Pressure is in psi, with the static gain or loss from height at 0.433 psi per foot, or read the other way as one psi for every 2.31 ft of head. Pipe is sized by nominal diameter in inches, and friction loss is given as psi per 100 ft of pipe. Velocity is in feet per second.

WSFU

Water supply fixture unit, a dimensionless load value for a fixture, folding flow, duration, and frequency of use

Probable demand

The diversified peak flow in gpm from the WSFU total read off the Hunter demand curve, not the sum of fixture flows

Diversity

The statistical fact that not all fixtures draw at once, which the demand curve builds in

Residual pressure

The flowing pressure left at a fixture with the water running, reserved before any friction is spent

Static head

The pressure to lift water, about 0.43 psi per foot of elevation, or 1 psi per 2.31 ft

Developed length

The measured pipe run along its actual path plus the equivalent length of fittings, not the straight-line distance

Flushometer vs flush tank

A valve flushing a closet directly in a short high-flow burst versus a slow-filling tank; the two read different demand lines and need different residual pressure

Erosion corrosion

Pipe wall and fitting wear from velocity above the limit, seen as grooves and pinholes at elbows

Related tools

Calculators and readiness checks for this work

Compare your options

FAQ

People also ask

• What slope does a drain need?
• How do you size a hot water recirculation pump?
• What causes water hammer?
• How do you size a water heater?
• How do you size a grease interceptor?