What a Wet Pipe and a Dry Pipe System Each Solve

Why this matters

Dry pipe systems get chosen for the wrong reason constantly, usually by someone reasoning that water in a pipe is a liability and air is not. The trade is real and it runs the other way from how it sounds: a dry system buys freeze protection and pays for it in delivery delay, in a larger water demand for the same protection, and in a worse corrosion regime than the wet system it replaced. A wet system puts water on the fire the instant the first head opens; a dry system has to trip a valve and blow the air out of every foot of pipe between the valve and that head first. There is exactly one condition that makes the dry system the right answer.

The gate

Can this pipe be kept above freezing, every hour of every winter, without depending on anything a tenant can turn off?

That is the whole decision. The adopted edition of NFPA 13 sets a minimum maintained temperature for areas containing wet pipe, commonly stated as 40 F, and that number is the gate. Above it, wet. Below it or unreliably at it, something other than wet. Every other argument for dry pipe (water damage worry, an insurer's preference stated without a reason, a previous building the owner remembers) is a reason to have a conversation, not to change the system type.

What a wet system is

Water stands in the piping to every head at supply pressure. A head opens, water leaves immediately, and a waterflow switch reports to the fire alarm panel. One moving assembly in the ordinary case, an alarm check valve or a simple riser check, and nothing to trip.

Four consequences worth naming. Zero delivery delay, because the water is already there. The simplest failure surface: the system is unavailable essentially only when a valve is shut, which is why the control valve tamper switch is the highest-value supervisory signal in the building. The least corrosive interior, because a pipe full of standing water consumes its dissolved oxygen once and then largely stops. And a leak is a wet leak: a damaged head discharges at supply pressure until the control valve is closed, and closing that valve starts an impairment.

What a dry system is, mechanically

The piping is filled with compressed air or nitrogen and held closed at a dry pipe valve. The valve is built so the air side has more clapper area than the water side, letting a modest air pressure hold back a much higher water pressure, a differential in the neighbourhood of five or six to one depending on the valve; the valve's data owns the ratio and the air pressure to maintain. When a head opens, air escapes, the differential collapses, the clapper trips, and water fills the pipe.

Three things exist only in dry systems, each a maintenance item with no wet-system equivalent:

• An air supply and its supervision. A compressor or nitrogen source with a pressure switch and a low-air supervisory signal at the panel. A low-air signal is not a nuisance, it says the differential is drifting toward an unintended trip.
• Pitch and auxiliary drains. Dry piping is pitched so condensate runs back to drain points, commonly at least a half inch per 10 ft on branch lines and a quarter inch per 10 ft on mains, with the adopted edition owning the figures. Trapped sections get an auxiliary drain, the two-valve arrangement often called a drum drip, drained seasonally or the trapped water freezes and blocks the pipe.
• Quick-opening devices. An accelerator or exhauster that dumps air faster so the valve trips sooner. The adopted edition caps system volume and sets a water delivery target, commonly 60 seconds to the inspector's test connection, and requires a quick-opening device or a demonstrated calculation above a volume threshold. Both figures belong to that edition, not to a rule of thumb.

The two costs a dry system charges

Cost one: delivery delay. Between the head opening and water arriving there is trip time plus transit through the pipe, and during that interval the fire keeps growing and more heads may open. Hence the second cost.

Cost two: a larger design area. The adopted edition of NFPA 13 requires the area of sprinkler operation to be increased, commonly by 30 percent at the same density, for dry pipe systems. That is not a penalty, it is the design acknowledging that more heads will be open by the time water arrives, and it flows straight into the water demand, which is density multiplied by area of operation, a relationship the sibling coverage card owns.

And a third cost that is not in the calculation: corrosion. The direction is the opposite of intuition. A dry pipe system holds residual water in low spots and traps, holds air above it, and the air supply keeps delivering fresh oxygen. That combination corrodes faster than a wet system, not slower, showing up as pinhole leaks, obstructed small orifices, and corrosion products migrating toward heads. It is live enough that nitrogen replaces compressed air on many dry systems specifically to take oxygen out of the equation.

The other two answers to the same gate

• Dry pendent or dry sidewall heads off a heated wet system. A sealed extension tube keeps water back at a heated branch line while the head sits in the cold space. Right answer for a handful of heads over a dock door, a canopy or a walk-in cooler. The tube is ordered to length and never cut or extended in the field.
• A listed antifreeze loop. Permitted in a narrower set of cases than it used to be, using listed premixed solutions, with limits on loop size and location. This has moved substantially across editions, so it routes to the adopted edition and the authority having jurisdiction rather than being decided from experience.

One gate, two spaces in the same building

A single-tenant distribution building in a cold climate. The owner asked for the whole building to be converted to dry pipe after a burst pipe in a neighbouring property. Two areas, and the gate answers them oppositely.

Space one: the exterior loading dock canopy, roughly 900 sq ft of covered area, four heads, open to outside air on three sides.

Gate answer: no. Nothing keeps this above 40 F in January, and the only thing that could is a heater a tenant can switch off, which is exactly what the gate excludes. Wet is off the table.

The choice is between dry pendent heads off the adjacent heated wet system and a small dry pipe system. Four heads is well inside the range where dry pendents are the cleaner answer: no dry valve, no air supply, no auxiliary drains, no delivery delay beyond the short tube, and no design area increase because the feeding system is still wet. Order four dry pendent heads against the measured distance from the heated branch line to the ceiling line at each location.

Space two: the main warehouse floor, ordinary hazard group 1, heated and held at 50 F, roughly 40,000 sq ft.

Gate answer: yes, this pipe stays above 40 F. The owner's instinct says "the warehouse is cold" and 50 F is above the gate, so wet is correct. Work the cost of converting anyway, because the owner deserves the number.

Take the design as an ordinary hazard density of 0.15 gpm per sq ft over a 1,500 sq ft area of operation, figures from the hazard classification and the design documents rather than from this card. Wet system demand is 0.15 times 1,500, which is 225 gpm, before the hose stream allowance that the coverage card owns and that is added identically to both options, so it is left out deliberately.

Converting to dry increases the area of operation by 30 percent at the same density: 1,500 times 1.30 is 1,950 sq ft, so demand becomes 0.15 times 1,950, which is 292.5 gpm. That is 67.5 gpm more, a 30 percent increase over the 225 gpm the wet system needed.

Corrections, printed.

• Raw comparison: dry against wet at the same density and area. Correction applied: the area of operation is re-based upward by 30 percent for a dry system rather than the density being increased, so this is a re-basing of the area, not an addition to the density, and every downstream figure uses the re-based 1,950 sq ft. Corrected demand: 292.5 gpm.
• Raw hose stream allowance: not shown. Correction applied: it is added at the point of connection and is identical for both options here, so it is excluded from this comparison by choice and named as excluded rather than quietly dropped.
• Raw temperature basis: "the warehouse is cold." Correction applied: the gate is the maintained temperature at the pipe, 50 F, against a 40 F floor from the adopted edition. Corrected reading: above the gate, so wet.
• Raw corrosion assumption: air is cleaner than water. Correction applied: reversed. A dry system's residual water plus a continuously refreshed oxygen supply corrodes faster than standing water.

The answer. The dock canopy gets four dry pendent heads and the warehouse floor stays wet. Converting the warehouse would have required 30 percent more water from a supply sized for 225 gpm plus hose allowance, added a delivery delay to every head, introduced an air supply and auxiliary drains to maintain, and made the corrosion regime worse, all to solve a problem the space does not have.

Sibling-rule check on both spaces. Water demand was computed as density times area of operation rather than from a head count, consistent with the coverage card. The 0.15 density came from the hazard classification and was not re-derived here, consistent with the classification card. The dry pendent heads were specified as a listed head type with an orientation and a tube length rather than a generic substitution, consistent with the head types card. No control valve was closed during the assessment, and the dock work is scheduled as a tagged impairment on the wet system with a compensating measure and a documented close-out.

Failure mode. The way this goes wrong is a conversion sold on comfort rather than on the gate. The building ends up with a system that is slower, thirstier and quietly rusting from the inside, and the first sign is a low-air supervisory signal acknowledged and never investigated, followed a year later by a pinhole leak in a pitched main. The second failure mode is smaller and more common: the auxiliary drains never get drained, a trapped section freezes, and a branch line stays blocked with ice all season with nothing at the panel to say so.

How to verify you got this right

1. Read the maintained temperature, do not infer it. Put a logger in the coldest corner through a winter week rather than asking what the thermostat is set to. The gate is 40 F at the pipe, and pipe near an exterior wall at a dock door is colder than the room average.
2. On any dry system, find every auxiliary drain and confirm it is on a drain schedule. A drum drip nobody knows about is the likeliest blocked section in the building.
3. Drain an auxiliary drain in the correct order and no other. Close the upper valve first, open the lower to drain the trapped section, then close the lower and reopen the upper. Opening the lower first on a charged system vents system air toward the trip point, and a dry valve that trips unintentionally floods the space and puts the system out of service.
4. Check the low-air supervisory history at the panel. Repeated events mean the air supply is chasing a leak, and a leaking dry system is one unattended weekend from an unintended trip.
5. Never break into a pressurized dry system. Close and tag the control valve under the impairment program in NFPA 25 in the edition your authority having jurisdiction has adopted, relieve system air through the valve's own vent, and confirm zero on both gauges before any fitting is loosened, because a dry system holds stored energy in the air volume as well as water behind the clapper.

References

• NFPA 13, in the edition adopted and amended by the authority having jurisdiction, for the minimum maintained temperature, pitch, water delivery time, system volume limits, the design area increase for dry systems and the rules on antifreeze loops
• NFPA 25, in the adopted edition, for dry valve trip testing, air pressure maintenance, auxiliary drain schedules and the impairment program
• Dry pipe valve and quick-opening device manufacturer data, which owns the differential ratio and required air pressure
• See related: Why Sprinkler Coverage Is a Density Problem, Not a Head Count; What a Preaction System Is Protecting Against