Fire Riser Spools: Trapped Air, Tee Losses & Test Ports

Three small choices decide how a fire riser behaves on test day: how the joints are made, where the air goes when you fill, and which way the water turns at a tee. All three can be settled with arithmetic before the first spool is lifted. Here are the sums we use at LEDE, with a few common figures corrected.

• A 5 L air pocket compressed for a 2 MPa test stores about twice the energy of a cubic metre of water at the same pressure
• A branch tee costs about three times the head loss of a run tee, which adds up fast at riser velocities
• A self-sealing test port still releases a small amount of water when the probe leaves, so "zero drip" needs a measured figure

Do Prefabricated Pipe Spools Really Beat Field Welding in a Riser?

On time and cleanliness, yes. On hydraulic loss from weld beads, less than most outlines say.

Start with the bead. A weld root that sticks 3 mm into a 154 mm bore narrows the passage to 148 mm. The loss from that ring is roughly (1/β² − 1)² of the velocity head, where β is the diameter ratio of 0.961. That works out near 0.007, which is a rounding error next to one tee. So the argument against an internal bead is not lost head. It is loose slag. A flake of slag or a spatter ball can break free during the first high-flow flush and travel to a dead end or a sprinkler. It is also a place where a crevice can corrode. We therefore treat debris, not friction, as the case for shop fabrication.

Then the time. Heat input for arc welding is Q = η × U × I ÷ v. Take 24 V, 150 A, an efficiency of 0.8 and a travel speed of 2.5 mm/s. That gives about 1.15 kJ/mm. A DN150 pipe has a 168 mm outside diameter and a 528 mm circumference. One pass at 2.5 mm/s takes about 211 seconds, and a root, fill and cap pass take over 10 minutes of arc time alone. Add fit-up, cleaning, chipping slag, cooling and a hot-work permit with a fire watch, and 45 to 60 minutes per joint is a common planning figure, though it varies by crew. A grooved joint takes two bolts and a gauge check. We use roughly 10 to 15 minutes as a planning number. The saving, close to the 75 percent some outlines quote, depends on your crew and site.

Groove dimensions come from AWWA C606, and the allowed variation depends on the pipe size and on whether the groove is cut or rolled. We therefore do not quote one tolerance for every size. Ask for the C606 table value for the size you buy, and ask how the shop checks groove diameter and depth with a gauge on each spool.

How Much Energy Does Trapped Air Store, and How Should a Bleed Ring Vent It?

More than most people expect, and the sum is short. Water is nearly incompressible. Air is not. Put 5 L of air at atmospheric pressure into a riser and bring the system up to 2 MPa gauge, about 2.1 MPa absolute. For adiabatic compression, with γ = 1.4, the pocket shrinks to 5 × (0.101 ÷ 2.101)^(1 ÷ 1.4) ≈ 0.57 L. The energy stored is (P₂V₂ − P₁V₁) ÷ (γ − 1) ≈ (1,200 − 505) ÷ 0.4 ≈ 1.7 kJ.

Compare that with the water. The energy stored by compressing a cubic metre of water to 2 MPa is ½ × β × P² × V, with β = 4.6 × 10⁻&sup9; per Pa. That is about 0.9 kJ. So 5 L of air holds about twice the energy of a whole cubic metre of water. If a joint lets go, the air expands and pushes water out like a piston. That is why we vent high points before we pressurise.

A bleed ring puts a tapped port between flange faces for that purpose. Sizing the port is a velocity sum. Air must leave at the rate water arrives, so vent area A = Q ÷ v. If you fill at 5 L/s and want air to leave at no more than 30 m/s, to avoid spitting and noise, you need about 167 mm², which is a hole of roughly 14.6 mm. One 12.7 mm port at 30 m/s passes only about 3.8 L/s. The practical rule is to fill slowly, or to use more than one vent. At 1 L/s one small port is enough.

The same logic helps with test ports. A self-sealing test port, often called a Pete's plug, takes a probe about 3.2 mm across. At 2 MPa the probe is pushed back out with a force of P × A = 2 × 10⁶ × 8.0 × 10⁻⁶ ≈ 16 N, so the operator holds it in. Withdraw a probe that has been inserted 10 mm and the hole it leaves has a volume of about 80 mm³, or 0.08 mL. A good elastomer closes behind the probe, but you can expect to wipe a fraction of a millilitre. We therefore ask for a measured leak figure, and we do not say that a port never drips. The rated pressure and temperature of the port belong on the drawing, taken from its datasheet.

Fire Riser Spool Comparison by Build

Property	LEDE Prefabricated Grooved Spool With Vent and Test Ports (Targets)	Field-Welded Flanged Carbon-Steel Header	Threaded Black Iron Run	Unverified Low-Cost Grooved Spool
Joint method	Mechanical coupling, no hot work	Arc welding with a hot-work permit	Threads, pipe dope	Mechanical coupling
Internal slag or bead	None from joints, shop-checked	2 to 4.5 mm beads are common	Thread burr, dope	Not stated
Groove or end control	AWWA C606 table value per size, gauged	Weld gap, set on site	Thread depth	Often not gauged
Air venting	Bleed ring at high points, sized to fill rate	Drilled and welded boss	Usually none	Usually none
In-service pressure check	Self-sealing test port, fraction of a mL on withdrawal	Gauge needs isolation and draining	Gauge needs isolation and draining	Varies
Tee loss data	K values by flow path, tested or from standard tables	Not usually given	Not usually given	Not given
Site labour per joint	About 10 to 15 minutes (planning figure)	About 45 to 60 minutes (planning figure)	About 20 to 30 minutes	About 10 to 15 minutes

"Most trouble on a riser comes from what the drawing does not show: a pocket of air, a flake of slag, a tee turned the wrong way. Ask where the air leaves, where the debris would come from, and which way the water turns. Then ask for the number that proves each answer."

— Guo Wei, Chief Metallurgical & Critical Infrastructure Piping Specialist, LEDE

What Is the Penalty of a Tee Branch Compared with a Tee Run?

Take a DN150 tee with a bore near 154 mm. Textbook values (for example, the Crane method with a pipe friction factor fT of about 0.015) give K ≈ 20 fT = 0.30 for flow straight through the run, and K ≈ 60 fT = 0.90 for flow turned into the branch. These are standard rule values and not LEDE test results. The branch costs three times as much. Equivalent length is Le = K × D ÷ f. With a pipe friction factor of about 0.017, the run is 2.7 m of pipe and the branch is 8.1 m.

Velocity decides how much that matters. At 30 L/s the flow speed in DN150 is 1.6 m/s. The velocity head is 0.13 m, so a branch tee costs about 0.12 m (1.2 kPa) and a run tee 0.04 m. Both are small. Raise the speed to 6 m/s, a common upper design limit, and the velocity pressure is 18 kPa. A branch tee then costs 16 kPa and a run tee 5.4 kPa. Six branch tees at that speed cost about 97 kPa, against 32 kPa for six run tees. The 65 kPa difference is 0.65 bar, and it comes straight off the pressure at the sprinklers at the end of the line.

So the tee is a design choice. Where the flow can continue along the main, use the run. Where a branch is needed, place it late, after the speed has dropped. Ask suppliers for K values by flow path and size. If no test data exists, use the standard tables and say so.

Three-step acceptance checklist for a fire riser spool:

1. Gauge the groove diameter and depth on a sample of spools against the AWWA C606 table value for that size, and look inside the bore for debris
2. Count the high points on the drawing, size the vents to Q ÷ v for your fill rate, and plan a slow fill
3. Probe a test port at the test pressure, measure what comes out when the probe is withdrawn, and record the tee flow path on each branch

ledefittings.com, LEDE grooved pipe fittings and valves, supplies prefabricated grooved pipe spools, bleed rings and grooved fittings for fire protection pipework. All figures here are design targets or worked examples unless a test report says otherwise; documentation is available on request.