Pump Sizing Calculator Guide

Centrifugal pump sizing adheres to ISO 5199 (rotodynamic pumps—technical specifications), ANSI/HI 1.3 (pump efficiency testing), and ISO 9906 (hydraulic performance acceptance tests). These standards define test methods, uncertainty limits (±2–3% for head/flow, ±5% for power per HI 40.6), and reporting requirements. For system design, ASME B31.4/B31.8 (liquid/gas pipelines) and EN 13724 (pump selection for building services) provide context-specific guidance. Our calculator applies the fundamental hydraulic power equation (P = ρgQH / (η × 1000)) per ISO 5199 Annex A, but final selection must comply with manufacturer-certified test data—not just theoretical outputs. Always validate against pump affinity laws and derate for viscosity effects per ISO/TR 17766 when fluid kinematic viscosity exceeds 20 cSt.

The calculator’s power output (kW) has a theoretical accuracy of ±1.5% under ideal conditions, as it uses the standard hydraulic power formula corrected for efficiency and density. However, field deviations commonly reach ±8–12% due to unmodeled factors: pipe friction losses not included in head input, suction-side NPSH margin errors, fluid temperature-induced density/viscosity shifts, impeller wear (reducing η by 5–15%), and motor-drive inefficiencies (VFD losses, bearing drag). Per ANSI/HI 9.6.1, actual installed power should be verified via torque metering or calibrated wattmeter—not calculated values alone. Always add a 10–15% safety margin on motor rating per NFPA 70 Article 430.6(A)(1) for continuous duty.

Yes—but fluid density and viscosity must be rigorously accounted for. The calculator accepts custom density (kg/m³), essential for glycol solutions (e.g., 50% ethylene glycol ≈ 1090 kg/m³ at 20°C) or crude oil (≈ 820–950 kg/m³). However, viscosity >20 cSt invalidates the standard centrifugal pump curves: per ISO/TR 17766, head and efficiency drop significantly, requiring correction factors (e.g., HI 40.6-2019 charts) and potential impeller trim. Vapor pressure also affects NPSHr—critical for hydrocarbons (e.g., gasoline at 25°C: Pv ≈ 55 kPa → NPSHr increases ~0.5 m). Always cross-check with manufacturer viscosity-correction curves and validate suction conditions using API RP 14E for hydrocarbon service.

The 'pump_size' output is a normalized sizing index (not physical dimensions) derived from specific speed (Ns = N√Q / H^0.75) and power, correlating to typical frame sizes per ISO 5199 and ANSI B73.1. For example, Ns ≈ 1,200–2,500 suggests a close-coupled end-suction pump (ANSI B73.1 Type OH2); Ns > 4,000 indicates a double-suction or multistage design (ISO 5199 Class II). This index guides initial selection—e.g., 'pump_size 3.2' maps to ANSI 3×4×13 frame (3" suction, 4" discharge, 13" impeller diameter). Always verify mechanical details (shaft seal type, bearing life per ISO 281, flange rating per ASME B16.5) against manufacturer catalogs—not the index alone.

NPSHr rises with temperature primarily due to increased fluid vapor pressure (Pv), which directly reduces available NPSHa. Our calculator estimates NPSHr using empirical correlations tied to flow coefficient and impeller geometry—but assumes constant Pv. In reality, water at 80°C has Pv ≈ 47.4 kPa (vs. 2.3 kPa at 20°C), increasing required NPSHr by ~4.5 m. Per Hydraulic Institute Standards (HI 9.6.1), maintain minimum NPSH margin ratio (NPSHa/NPSHr) ≥ 1.3 for stable operation; ≥ 1.5 for high-speed or high-energy pumps. Always calculate Pv using Antoine equation or NIST REFPROP, and verify suction piping design (velocity < 2 m/s, no pockets) per ASME B31.1 to avoid localized boiling.

Yes—the efficiency input reflects pump (not motor) efficiency, typically 55–85% for centrifugal pumps. Realistic values depend on specific speed, flow, and head: per HI 9.1, best-efficiency-point (BEP) efficiency for Q=100 m³/h, H=50 m is ~72–78% for standard end-suction pumps (Ns ≈ 1,800). Use manufacturer published curves—not generic tables—as efficiency drops sharply off BEP (e.g., ±20% flow change reduces η by 10–15%). For abrasive or viscous fluids, derate by 5–12% per HI 9.6.5. Never assume 100% efficiency: ISO 5199 mandates reporting efficiency at rated point with ±3% tolerance. Field measurements via calorimetric or torque methods (per ISO 5198) are preferred for critical systems.

No—it calculates single-pump duty point at specified Q and H, assuming those represent the system’s static head plus friction loss at that flow. It does not model dynamic system curves, valve throttling (which increases head requirement without changing Q), or parallel/pump staging effects. For control valves, add their pressure drop to the input head; for parallel pumps, ensure each shares flow equally (per ANSI/HI 9.6.3, imbalance >10% causes recirculation damage). Use the calculator only for initial sizing—then simulate full system behavior with tools like AFT Fathom or PIPE-FLO, validating against Darcy-Weisbach friction (ASCE 78-18) and transient analysis per API RP 1142 for startup/shutdown.