Pump Sizing Calculation: A Step-by-Step Duty-Point Worksheet

Pump sizing calculation is based on the water a process needs and the resistance of the route. To size a pump, put those inputs and then pick horsepower. The worksheet below constructs a clean water duty point, shows all assumed losses, and segregates the calculated demand from the supplier evaluations prior to buying.

Step 1: Define what your pump sizing calculation must deliver

Pump duty combines a specific flow rate with the differential head at that flow. To begin, denote the liquid, the process demand and the two points at which energy is added. Describe the calculation condition with the operating requirements noted alongside the inputs.

For a transfer job, identify the source tank, the receiving tank, the fill volume and the fill time. For a process that’s continuously supplied, use the actual demand pattern. A pump feeding multiple units needs a credible simultaneous-demand case; adding every connected outlet at full demand may describe a condition the plant never experiences.

The U.S. Department of Energy’s pump selection advice separates properties of fluids from the end-use. Adhere to that. Before using water data, note the temperature, density, viscosity and any solids. Define the type of installation and proposed control, duty hours and the person responsible for verifying each input.

Keep the scope visible: The following instructions apply to one centrifugal pump transfer of clean water in steady-state conditions. Applications involving slurry, highly viscous liquids, the fire service, and transient pressure require their respective reviews. Two pumps operating in tandem also necessitate curves; the single-pump result below cannot establish their combined output.

The description of the 30 m³/h transfer indicates that the output will be a flow-and-head analysis with traceable assumptions. Motor nameplates can’t provide the missing pipe routing, tank levels, or suction data. Purchasing can request equivalent curves only after engineering defines that common duty and operations confirms that it actually occurs.

Step 2: Calculate the required flow rate

Moving 15 m³ in 30 minutes with no other concurrent demand requires a volumetric flow rate of 30 m³/h. Divide the required volume by the available interval time using matching units. Then determine if minimum, normal, and peak demands sufficiently differ to warrant separate operating scenarios.

Q = volume ÷ time = 15 m³ ÷ 0.5 h = 30 m³/h. For the head and power equations, convert the result to 30 ÷ 3,600 = 0.008333 m³/s. Keep full precision in the calculation and round the result only.

When considering the volume, the time period is equally important. A half hour isn’t a half an hour of pumping if a documented cleaning or sequencing activity is performed during that time. For the usable pumping time, ask operations to determine the time and calculate from the time confirmed by operations. Don’t include an unexplained flow percentage.

An entry should be made for minimum demand. Pumps that must supply one user overnight and several users during production may need different control arrangements from fixed batch-transfer pumps. Rather than guessing the normal duty for each shift, record the intended valve positions or speed-control setpoints.

To move 15 m³ within 0.5 h, the pumping window requires a flow rate of 30 m³/h on average. This arithmetic establishes the target; it doesn’t anticipate the actual flow rate of the installed pump. At the operating condition, final flow may differ due to system resistance, performance curve, and active controls.

Step 3: Build the total head calculation

Total differential head accounts for changes in elevation, pressure, velocity and velocity head losses along the flow path. Define the endpoints before using the equation. At large open tank free surfaces, pressures cancel due to atmospheric pressure, and surface velocities are usually negligible; however, velocity at the endpoints of a flowing pipe requires a different treatment.

H = (z₂ − z₁) + (p₂ − p₁)/(ρg) + (V₂² − V₁²)/(2g) + h_loss. Here H and elevation z are in metres, pressure p is in pascals, density ρ is in kg/m³, gravity g is in m/s², and endpoint velocity V is in m/s. The velocity term uses the usual kinetic energy correction approximation; however, a detailed energy balance must be performed to account for the actual flow profile if it’s relevant.

Use the same pressure basis at both endpoints. Two gauge readings corrected to the same atmosphere can comprise a pressure difference. For suction calculations involving vapor pressure, absolute pressure is required. Pressurizing the receiving vessel changes the pressure term even if its elevation doesn’t change.

For straight-pipe and local losses, the Hydraulic Institute system-curve method can be used: h_major = f(L/D)v²/(2g), and h_minor = ΣK·v²/(2g). Lowercase v refers to pipe velocity used for the loss term. Recalculate velocities for different pipe bores; don’t assume that the velocity of one segment is applicable to all fittings.

The Darcy friction factor, f, is dependent on Reynolds number and relative roughness. Darcy and Fanning factors are different conventions, so ensure your worksheet states which one they’re using. For the constant diameter example, the f and fitting coefficients are assumed to be fixed to simplify the process. An installed system would need to have justified roughness, fluid properties and actual internal diameters.

An enclosed, completely closed circulation loop wouldn’t require the pump in order to cover the entire height of the building on every pass. Flow losses and any differential pressure remain. Starting, filling and venting conditions would need to be addressed separately. For the pipe-loss portion, a separate reference is available in the friction loss formula guide.

With an 18 m difference between open free surfaces, the example begins with 18 m of elevation head and adds pipe losses at the selected flow. Adding a second arbitrary velocity-head allowance would change the result without changing the endpoints. Physically locate each term and then consider if it should be included in the balance.

Step 4: Follow the Duty-Point Evidence Sheet

The worked duty point is 30 m³/h at 22.97 m total head for the assumptions in the table below. All values are illustrated, including the bore of the pipe and loss coefficients. Replace all assumptions with verified site inputs before using this sheet to request a pump selection or to evaluate supplier proposals. The Hydraulic Institute system-curve tutorial explains the loss method; the numerical inputs below remain assumptions for this example.

The assumed K inventory is explicit: entrance 0.5, exit 1.0, three elbows at 0.5 each, check valve 4.0 and isolation valve 1.0, totalling 8.0. These are teaching inputs and not recommended lookup values for a particular fitting. These should be replaced with data that are geometry and position specific. The exit loss includes the discharge to the receiving tank.

Analysis at fixed flow indicates what the assumed friction factor changes. With flow set at 30 m³/h and all other inputs fixed, f = 0.018 results in a total head of 22.27 m and f = 0.028 yields 24.02 m. This isn’t a validated uncertainty band, but indicates a value requiring evidence.

The 22.97 m result can be calculated from the stated 18 m elevation, 3.85 m straight-pipe loss and 1.12 m local loss. Agreement between the Spreadsheets confirms the arithmetic only. When the evidence sheet reflects actual dimensions, coefficients and operating levels, it can support a real selection.

Step 5: Match the pump curve to the system curve

The operating point of a typical centrifugal pump is located where its performance curve satisfies the complete system requirement. Plot the actual supplier curve at the proposed speed and impeller diameter vs the system requirements. A calculated duty point alone can’t indicate whether the pump reaches this point or is located within the permitted operating region.

For this specific case, holding f and K constant yields H(Q) = 18 + 4.97310 × (Q/30)², with Q in m³/h and H in meters. The system requirement points are at 20.21 m at 20 m³/h, 22.97 m at 30 m³/h, and 26.84 m at 40 m³/h. These three points form a system curve on a head-versus-flow graph; they aren’t three certain pump outputs.

Along with the head curve, ask for pump efficiency and the manufacturer’s preferred and allowable operating regions. The Hydraulic Institute pump-curve guide cites that operation away from the best efficiency point can cause hydraulic and mechanical issues. Instead of a universal percentage, use the specific limits of that pump.

Active control valves require careful planning. A base system curve will usually not include the control valve loss. The controlled duty is then positioned above the base curve by the valve’s required pressure loss. Show the control requirement or make sure to show this discrepancy based on the combined pump and system curve. Never consider the base curve intersection as the controlled duty without accounting for control-valve loss.

Common failure modes: when not to order from one point

Don’t give an order when the supplier only sends a rated flow and head with no applicable curve. A point can align with the worksheet while the low-resistance case can drive the operation toward an excessive, high-flow boundary. Ask for shut-off head, the runout boundary, the maximum permissible suction and discharge pressures, and casing pressure limits. Transient pressure requires a separate consideration; steady-state head arithmetic can’t compute it.

The other boundary is important too. Zero or very low flow can cause pump damage, and this worksheet won’t define a safe deadhead period. Request the manufacturer’s minimum flow and protection specifications for the proposed pump and service. Explain how the installation will limit operation to those specified flows.

When two pumps are parallel, require a single curve for the composite system and individual operating checks for each pump for every desired stage of operation. Two identical pumps connected in parallel won’t automatically provide twice the flow of a single pump into the same resistant system. Arrangements in series need to be examined for combined head and pressure. Indicate whether the second unit is standby only or operating; those are different scenarios for engineering and purchasing.

Before changing speed to reach another point, consult the pump affinity laws guide to change speed for reaching another point and find the limits the supplier has provisioned.

Step 6: Check suction conditions before selection

Suction acceptance involves the available net positive suction head under the worst plausible conditions satisfying the pump’s requirement with a suitable application specific margin. Calculate the available suction head from absolute suction conditions and vapor-pressure head. Obtain the required suction from the manufacturer for the relevant flow, speed and liquid state.

Note the lowest source level, site atmospheric pressure, liquid temperature, friction and pressure losses in the suction line, and pump reference level. Available net positive suction head, NPSHA, signifies the total absolute suction head at the pump reference level less vapor-pressure head. Required net positive suction head, NPSHR, is a pump attribute. They aren’t interchangeable inputs, and the head calculated between two tanks doesn’t supply either value in itself.

In their March 2025 justification of the 2024 ANSI/HI 9.6.1 revision, Hydraulic Institute states a shift to the manufacturer provided NPSHR as the margin of reference. NPSHR must be equal to or greater than the tested value of NPSH3. Do not substitute an application review with a generic, older percentage from a tip sheet.

“The revised guideline provides users with NPSH margin recommendations specific to each application.”

Hydraulic Institute, explanation by Alex Moser, March 18, 2025

Intake geometry is a separate consideration. An adequate NPSHA doesn’t guarantee an open tank inlet has adequate submergence to prevent drawing an air-core vortex. This requirement is addressed in Hydraulic Institute’s submergence recommendations. For the review, please provide intake dimensions and lowest liquid level; no generic clearance is assumed.

Regarding the example of 30 m³/h, there’s no verified source for the relative level of the pump, the suction-line details or the range of liquid temperatures. Hence, the NPSHA and margin are undefined. Thus, carry the missing values along with the example, rather than accepting suction as indicated, because the complete head arithmetic for the other inputs is done. Procurement is to note and hold this as an open condition for the review of the quote.

Step 7: Calculate power at the correct boundary

Hydraulic power is the rate of energy transferred to the liquid, while shaft input and electrical input include different losses. At the illustrative duty, liquid power is 1.87 kW. Assuming 70% pump efficiency and 90% motor efficiency gives 2.68 kW at the shaft and 2.98 kW electrical input, before any drive losses.

In SI, P_hydraulic = ρgQH/1,000 = 998 × 9.81 × 0.008333333 × 22.97309948 ÷ 1,000 = 1.8743 kW. Divide by the assumed pump efficiency: P shaft = 1.8743/0.70 = 2.6776 kW. Divide again by the assumed motor efficiency: P electric = 2.6776/0.90 = 2.9751 kW.

These efficiencies are separate boundaries, not two approximations of the same loss. The Department of Energy centrifugal-pump performance sheet differentiates pump input and electricity consumption. For more detail of the equations, see the pump power formula and pump efficiency formula guides.

The 2.68 kW duty-point result alone cannot select a motor. The maximum absorbed power should be evaluated across the boundaries of all permitted credible operating states under the conditions of the speed, density of the liquid, and high flow. The motor and drive assessment requires applicable derating, starting requirements and manufacturer limits. A larger nameplate doesn’t correct an unacceptable range of operation for hydraulics.

Finance can use the electrical boundary for a clearly labeled running-cost figure. At an assumed 4,000 h/year and $0.12/kWh, the unrounded 2.9751 kW input yields approximately 11,900 kWh/year, or $1,428/year. A variable-duty installation requires hours and electrical input for each state. No purchase-price or maintenance comparison is supplied here, so this isn’t a payback claim.

A 2.68 kW shaft calculation indicates one assumed duty, not the maximum load a selected pump will impose. The illustrative annual electricity cost of $1,428 also depends on the stated hours, tariff and efficiencies. Ask each supplier to quote against the same operating schedule before finance compares energy costs between offers.

Step 8: Hand over the worksheet to a pump sizing tool and supplier

A useful sizing handoff identifies the normal duty, credible extremes and the evidence still needed for selection. Give operations, engineering and procurement the same set of inputs, then ask each supplier to return comparable curves and limits. A model-matching tool can narrow candidates after flow and head are established; it can’t verify missing field data.

Copy the following matrix into the enquiry and fill its site-condition values. The example provides the standard numerical target only. Don’t transform the three points on a single system curve into three confirmed plant operating cases.

Selection Bottleneck Map

Resolve the input that can change the selection prior to dealing around the nominal result. The owner column keeps an open engineering query from being an automatic buying allowance.

Digital tools help make those assumptions visible. The Hydraulic Institute’s December 2024 data Tool announcement describes some curve and calculation demonstrations and resources. This doesn’t signify the software can validate a request. Save the input version and the supplier’s returned curve.

Once inputs for relevant agricultural product families are added to BBP’s pump sizing calculator, use model-based matching. Include the evidence sheet and operating cases in the enquiry so that relevant application boundaries can be addressed in the proposal that’s returned.

FAQ: Pump sizing questions