← Fire Protection Systems
FIRE PUMP SERIESNFPA 20 §4.20NFPA 25 §8.3.3

Test Header
The Flow Test Manifold

The array of 2½-inch outlets that lets you dump a fire pump's rated flow safely onto the parking lot — and prove the pump still makes curve. Here's how to read the manifold in the field, why it lives on the exterior wall next to the FACP and fire riser room, how many outlets your pump actually needs, and the common installation mistakes that cost you during the annual flow test.

By Stanislav Samek, Samektra · 12 min read · Last updated May 3, 2026(3w ago)

Three 2½" outlets means this header is sized for a pump up to ~500 gpm rated (3 × 250 gpm = 750 gpm = 150% of 500 gpm rated). For larger pumps, more outlets — see the diagram below.

Standard service-wall configuration: everything life-safety-critical lives behind three supervised doors in one clearly-labeled 20-foot stretch of facade. This layout exists so that arriving fire marshals, flow-test contractors, and responding engines can orient themselves in seconds, not minutes.

What each labeled element does

Test Header Manifold System — multi-valve assembly for pump flow data
Fire Pump Test Header Valve — isolation valve upstream of the manifold
Test Port Discharge Coupling — where hoses attach (2½" NH threaded)
Manifold Wall-Mount Bracket — structural support bolted to the building
Fire Pump Test Header Label — NFPA 20 required identification
Supervisory / Flow Alarm Bell — tied to the FACP, audible outdoors during alarm
Fire Riser Room — supervised access to the sprinkler riser and alarm check valve
FACP Room — fire alarm control panel behind supervised door
Electrical Room — disconnect access for fire pump and controller
Facade Service Lights — illuminate the service area for night operations
Reinforced Concrete Sidewalk — graded for water discharge during flow test

Why Pumps Need a Test Header

NFPA 25 §8.3.3 requires an annual performance test of every fire pump: move rated flow through the pump at rated pressure, then 150 percent of rated flow at not less than 65 percent of rated pressure, and verify the churn (no-flow) pressure against the original acceptance curve. To do that, you need somewhere to dump thousands of gallons per minute.

A test header is a short branch piped off the fire pump discharge, typically exiting the pump room to an exterior wall, terminating in a row of 2½-inch hose valves mounted in a weather-protected enclosure. During the test, inspectors attach hoses, open valves one at a time, and use pitot tubes on short playpipes to measure flow out of each hose. Multiple outlets are needed because each 2½-inch outlet only carries about 250 gpm at standard test conditions — the NFPA design convention.

How Many Outlets Do You Actually Need?

Six outlets, six hose connections, one manifold body. This is what “enough capacity” looks like for a typical mid-rise commercial fire pump.

The 250-gpm-per-outlet convention

This is the single most important number for sizing a test header: each 2½" outlet is designed for 250 gpm. That figure comes from the NFPA 14 standard fire-stream basis — a 1⅛" smoothbore nozzle flowing at approximately 80 psi nozzle pressure, the same standard used for fire department attack hose. It's conservative by design: real-world flow through a single outlet can reach 325–500 gpm with a 1¼" or 1⅜" tip at higher nozzle pressures, but the code-compliant design assumption is 250 gpm per outlet, flat.

This means sizing is simple arithmetic: outlets needed = (pump rated flow × 1.5) ÷ 250, rounded up. Undersized headers are a common legacy problem — a pump upgraded from 1,000 to 1,500 gpm without a matching test header upgrade leaves you physically unable to perform the annual test without workarounds.

Pump rated flow150% test flow requiredOutlets @ 250 gpm eachTypical install
250 gpm375 gpm2 outlets3 (spare)
500 gpm750 gpm3 outlets3
750 gpm1,125 gpm5 outlets5
1,000 gpm1,500 gpm6 outlets6
1,500 gpm2,250 gpm9 outlets9 or flow meter combo
2,000 gpm3,000 gpm12 outletsFlow meter loop preferred
2,500+ gpm3,750+ gpm15+ outletsFlow meter + hydrant dump combo

💡 Why 250 gpm — the hose-stream math

The discharge formula Q = 29.84 · c · d² · √P with a 1⅛" tip (d = 1.125), a smoothbore coefficient c ≈ 0.98, and nozzle pressure P = 80 psi gives Q ≈ 330 gpm theoretical. Real-world losses (hose friction over the short test length, coupling turbulence, gauge calibration) bring the practical flow to the 250 gpm design value codified in NFPA design practice. Higher nozzle pressures (up to ~100 psi) with the same 1⅛" tip push 370+ gpm, and larger tips (1¼" or 1⅜") push 425–500 gpm — but those are real-world maxima, not design figures. Always size the test header to 250 gpm per outlet for code compliance.

⚠️ The undersized-header problem

If your test header can't physically discharge 150% of rated flow at 250 gpm per outlet, NFPA 25 §8.3.7 forces you into an alternative: flow meter loop (permanent install), hydrant dump (AHJ approval + water utility coordination), or supplemental discharge. All three add cost and complexity. The photo above shows a 3-outlet test header — sufficient for a 500-gpm pump but clearly undersized for anything larger. If you're doing any pump room renovation, bring this up during the design phase — adding two more 2½" valves at install is cheap; retrofitting later is not.

How to Read the Flow

Each open hose valve is measured with a calibrated pitot gauge placed in the center of the stream just outside the smooth-bore nozzle. The pressure reading (in psi) is converted to flow (in gpm) using the nozzle coefficient and discharge formula: Q = 29.84 · c · d² · √P, where c is the nozzle coefficient (0.90–0.97 typical), d is the nozzle diameter in inches, and P is the pitot pressure in psi. Add up each individual flow to get the total discharge, then plot it on the acceptance curve.

Three required test points per NFPA 25 §8.3.3:

  1. Churn (no flow): all test header valves closed. Pump produces shutoff pressure. Must match the original acceptance curve within tolerance.
  2. 100% rated flow: open enough outlets to discharge the pump's rated flow (e.g., 1,000 gpm on a 1,000-gpm pump). Pump must maintain at least rated pressure.
  3. 150% rated flow: open additional outlets to reach 1.5× rated flow. Pump must maintain at least 65% of rated pressure. This is where most degraded pumps fail.

Common Field Issues

Missing or frozen caps

Caps protect outlets from debris and vandals but are themselves prone to theft, weather damage, and winter ice. A frozen cap can delay testing for hours while it thaws. Replace missing caps immediately; use chained caps to prevent loss.

Corroded or painted-shut valves

Outdoor-mounted valves corrode over decades. A painted-shut valve handle looks fine on quarterly inspection but will not rotate during flow test. Exercise each valve through full travel annually per NFPA 25 §8.3.3.1.

Missing or degraded gaskets

Rubber gaskets at the hose coupling dry-rot, crack, or go missing. Leaky couplings lose measurable pressure during the test and produce incorrect flow data. Keep spare gaskets on-site.

Undersized manifold

Not enough outlets for 150% flow. Pump upgraded without test header upgrade is the typical cause. Forces use of NFPA 25 §8.3.7 alternatives (flow meter, hydrant dump).

Poor drainage grading

Concrete service sidewalk not graded away from the building, or to a storm drain. Water pools or backflows against the foundation. Can also flood neighboring property — complaint risk.

Missing identifying signage

NFPA 20 requires a FIRE PUMP TEST HEADER label. Without it, arriving technicians may mistake the manifold for an FDC, or fire marshals may cite the unlabeled assembly.

Cross-connection risk

Header piped upstream of the backflow preventer creates a cross-connection to the potable water supply. Immediate correction required by water utility — not something to leave until the next inspection.

Obstructed access

Parked vehicles, dumpsters, landscaping, or construction materials blocking the test header. Same principle as FDC access — if it can't be reached quickly, it may as well not exist.

Safety During Flow Testing

A 1,500-gpm flow test through multiple 2½" hose lines is not a trivial operation. Reaction forces at each nozzle can exceed 50 lbs; unsecured hoses whip; and the discharge area is a wet, slippery, high-pressure environment. Best practices:

  • Secure each hose. Two-person minimum per active line, or use hose restraints / playpipe stands. Never attempt to hold a single 2½" line solo at rated pump pressure.
  • Block vehicle and pedestrian access to the service area during the test. Close the fire lane, post cones, and have a flagger if the area fronts a public walkway. The photo above shows proper safety cones in place — and what it looks like when one falls over.
  • Verify drainage capacity before opening valves. A 1,500-gpm dump into an under-sized storm drain backs up and floods the site. Know the drain capacity and grade.
  • PPE: hearing protection (pumps and flow noise), safety glasses, gloves, non-slip footwear. Hi-vis vest if traffic is present.
  • Radio contact between the pump room operator and the outdoor test crew. Signals: open X valves, hold, close X valves, back to churn. Pitot readings are called out by line and written down — no shouting over running water.

Alternatives — Flow Meters and Hydrants

Where a test header is impractical (roof-mounted pumps, urban sites with no drainage, oversized pumps with insufficient outlets), NFPA 25 §8.3.7 allows alternatives:

  • Flow meter on a test loop: water recirculates back to the suction supply through a calibrated meter. No water leaves the building. Accurate and no drainage concerns, but $30,000–$80,000 to install.
  • Hydrant dump: discharge into a nearby hydrant-flow drain or storm system. Requires AHJ approval and coordination with the water utility to avoid brown-water complaints and to notify downstream pressure-sensitive customers.
  • Suction-side flow measurement: rarely used because it requires a meter on the supply line that is always in service — a failure mode in the meter = no fire water.
  • Combination approaches: test header for the 100% flow point, then flow meter or hydrant dump to add the additional flow needed to reach 150%. Workable when the test header is slightly undersized.

How to Read the Flow

Each open hose valve is measured with a calibrated pitot gauge placed in the center of the stream just outside the smooth-bore nozzle. The pressure reading (in psi) is converted to flow (in gpm) using the nozzle coefficient and discharge formula: Q = 29.84 · c · d² · √P. Add up each individual flow to get the total discharge, then plot it on the acceptance curve.

Alternatives — Flow Meters and Hydrants

Where a test header is impractical (roof-mounted pumps, urban sites with no drainage), NFPA 25 §8.3.7 allows alternatives:

  • Flow meter on a test loop: water recirculates back to the suction supply through a calibrated meter. No water leaves the building. Accurate but expensive to install.
  • Hydrant dump: discharge into a nearby hydrant-flow drain. Requires AHJ approval and coordination with the water utility to avoid brown-water complaints.
  • Suction side flow measurement: rarely used because it requires a meter on the supply line that is always in service.

▶ Watch: Fire pump test header — field example

Source: Field walkthrough · Open on YouTube ↗

Frequently Asked Questions

What is a fire pump test header?
A test header is a short branch of piping off the fire pump discharge that terminates in a row of 2½-inch hose valves on the exterior of the building. During the annual NFPA 25 §8.3.3 performance test, inspectors connect hoses to each valve and discharge water through pitot-equipped nozzles. Individual flows are measured, then summed, to prove the pump delivers rated flow at rated pressure and 150% of rated flow at ≥65% of rated pressure.
How many outlets does a test header need?
NFPA design convention: each 2½" outlet is sized at 250 gpm (a 1⅛" smoothbore nozzle at ~80 psi nozzle pressure — the same standard used for fire department attack hose). So a test header needs enough outlets to discharge 150% of rated flow at 250 gpm each: 500 gpm pump = 3 outlets (750 gpm / 250), 1,000 gpm pump = 6 outlets (1,500 / 250), 1,500 gpm pump = 9 outlets (2,250 / 250), 2,000+ gpm pump = 12+ outlets (typically switched to flow meter loop or combo). Some test headers use higher-flow configurations (1¼" or 1⅜" tips at higher nozzle pressures, pushing 325–500 gpm per outlet), but the NFPA default for sizing is the conservative 250 gpm figure.
Why is the test header on an exterior wall?
Three practical reasons: (1) the ~2,000+ gpm of water flowing during the 150% test has to go somewhere — typically a paved service area graded away from the building, not inside the pump room. (2) Inspectors need clear space to connect hoses, operate pitot tubes, and observe flow streams safely. (3) Co-locating with the fire riser room, fire pump test header label, FACP, and electrical room keeps all life-safety access points in one service area that emergency responders and inspectors can find quickly. Labeling is critical — per the photo above, the "FIRE PUMP TEST HEADER" sign is required per NFPA 20 so arriving technicians (and fire marshals) identify the manifold at a glance.
What does each component in the test header manifold do?
Standard components on a multi-valve test header: (1) Isolation valve — shuts off the entire header from the fire pump discharge when not in use. (2) Test header manifold — the horizontal pipe run with the outlets. (3) Individual 2½-inch hose valves — one per outlet, each independently opened during the test. (4) Test port discharge couplings — where hoses attach (NH threaded or Storz). (5) Wall-mount bracket — structural support for the manifold. (6) Drain / ball drip — prevents trapped water from freezing in winter. Optional: pressure gauge tap for back-pressure verification during flow.
What are the flow test requirements?
Per NFPA 25 §8.3.3: (1) Churn (no flow) — pump must produce rated pressure +/- tolerance against the original acceptance curve. (2) 100% rated flow — pump must produce at least rated pressure (tolerance per §8.3.3.3). (3) 150% rated flow — pump must produce at least 65% of rated pressure. All three points must be plotted and compared to the original pump curve. If any point falls outside the acceptance envelope, the pump is considered degraded and corrective action is required.
What alternatives exist when a test header cannot be used?
NFPA 25 §8.3.7 allows: (1) Flow meter on a test loop — water recirculates back to suction through a calibrated meter. No discharge to atmosphere, accurate, but expensive to install. (2) Hydrant dump — discharge into a nearby hydrant-drain arrangement or storm drain. Requires AHJ approval and coordination with the water utility. (3) Suction-side flow measurement — rarely used, requires permanent meter on the supply line. Every installation should have a primary test method specified on the pump acceptance paperwork.
What are common test header field problems?
Seven frequent findings: (1) Missing or frozen caps — cannot be removed for testing. (2) Corroded or painted-shut hose valves. (3) Gaskets missing or degraded, causing leaks during test. (4) Undersized manifold — not enough outlets for 150% flow. (5) Drainage not graded away from building — water pools or floods neighboring property. (6) No identifying signage — technicians and fire marshals can't identify it. (7) Cross-connection risk — header piped upstream of the backflow preventer, requiring immediate correction. Quarterly visual inspection catches the first three; an annual walkthrough before the flow test catches the rest.

References

1. NFPA 20 (2022), §4.20 — Fittings and test arrangements for flow measurement.

2. NFPA 25 (2023), §8.3.3 — Annual fire pump performance test.

3. NFPA 25 (2023), §8.3.7 — Alternatives when a test header is unavailable (flow meter bypass, hydrant dump, etc.).

4. NFPA 13 (2022), §16.9.7 — Test connection arrangements in combined systems.

5. IFC 2024: International Fire Code, §913 — Fire pumps and related equipment.

DISCUSSION
Be the first to contribute.

Open the discussion panel to comment, flag an inaccuracy, add field experience, or ask a question. Approved contributions earn SRP and may be incorporated into the article.