10.1 Fire Pump Types (Horizontal Split Case, Vertical Turbine, End Suction) & System Arrangements

Understand the pump before testing it

NFPA 20 (2022) addresses stationary fire pump installation; NFPA 25 (2023), Chapter 8, addresses ongoing ITM. An inspector needs the installation records to understand the equipment but must distinguish a routine condition assessment from a new installation design review. Record the pump model, serial number, rated flow, rated net pressure, rated speed, driver, supply, and original acceptance curve. Those facts establish the basis for later tests and prevent comparison with the wrong pump.

Most familiar building fire pumps are centrifugal. They transfer energy through a rotating impeller and convert part of that energy to pressure in the casing or diffuser. Positive-displacement fire pumps also exist, so “all fire pumps are centrifugal” is false. Identify the actual category before applying centrifugal performance expectations or test methods.

Common centrifugal configurations

A horizontal split-case pump has a casing that separates for access to its rotating assembly. A common double-suction arrangement feeds both sides of the impeller. It is useful to recognize the casing split and bearing arrangement, but not every pump with a horizontal shaft is identical. Servicing an upper casing still requires isolation, drainage, stored-energy control, and the manufacturer's lifting procedure. Accessible construction does not permit opening a running or pressurized pump.

An end-suction pump receives water through an axial inlet at the end of its casing. It commonly uses an overhung impeller and a more compact layout. A vertical turbine pump has submerged bowl assemblies connected through a column and line shaft to the surface driver and discharge head. It is suited to arrangements where the water surface is below the driver elevation. The bowls must have the required submergence at the relevant pumping condition; a universal “second impeller” rule is not a substitute for the actual design.

Read the characteristic curve

For the usual centrifugal fire pump performance envelope, rated capacity is delivered at rated net pressure, 150% rated capacity is delivered at not less than 65% rated net pressure, and shutoff pressure does not exceed 140% rated pressure. There is no universal 101% minimum churn criterion. The actual certified curve and the NFPA 25 condition-evaluation criteria matter; the broad installation envelope alone does not establish that a used pump is performing acceptably.

For a hypothetical 1,000 gpm pump rated at 100 psi net, the benchmark flows are 0, 1,000, and 1,500 gpm. The 150% installation-envelope pressure is at least 65 psi net. These are net pump pressures, not automatically discharge-gauge pressures. A 45 psi suction supply adds to the discharge pressure, subject to the appropriate elevation and velocity corrections. Confusing gross discharge pressure with net pump pressure can hide pump deterioration.

At comparable operating points for the same centrifugal pump, flow varies approximately with speed, head with speed squared, and power with speed cubed. These affinity relationships help normalize valid test results within their applicable limits. They are not permission to increase engine speed beyond its rated setting to make a failing test look acceptable. Variable-speed and other special arrangements require their specific procedures.

Recognize supply and discharge components

Trace the suction source and control valves, pump, discharge check valve, discharge control valve, bypass where provided, and test connections. Suction restrictions, air pockets, damaged restraints, or an incorrect valve position can reduce available water before the pump ever develops its expected curve. An eccentric reducer in a horizontal suction arrangement is commonly installed to avoid an air pocket; judge the actual arrangement under the applicable NFPA 20 requirements and records.

Do not treat a quick visual survey as proof of every concealed installation dimension. Record apparent changes or suspect fittings for qualified evaluation. Check that required valves are in their prescribed normal position and supervised or secured as applicable. A test-header isolation valve is normally closed; the system discharge control valve is normally open. These valves serve different purposes and must not be confused during a churn test.

Distinguish the small circulation-relief function that prevents overheating at churn from a main pressure-relief function. They are not interchangeable. Cooling arrangements vary with the driver and installation. Observe actual discharge, temperature, and operation under the prescribed test; do not assume a diesel heat exchanger makes every separate cooling feature unnecessary.

The test header or flow-meter arrangement must support the required test flows. Header valve counts come from the applicable NFPA 20 table and arrangement, not a universal “one outlet per 250 gpm” formula. Record instrumentation and test boundaries so a later reviewer can determine whether the external supply was actually exercised. Finish by checking leakage, supports, guards, valve status, and any limitation affecting readiness.

Sources checked October 19, 2026: NFPA 20 (2022), pump arrangements and performance provisions; NFPA 25 (2023), Chapter 8.