What are you working on?

Find a calculator, explore a formula, or learn the method.

THE ENGINEERING REFERENCE / GUIDE

Correcting Water-Based Pump Performance for Viscous Liquids: Required Data

Short Answer

Correcting a water-based pump performance curve for a viscous liquid requires four core inputs: the manufacturer's water performance curve with its best efficiency point values, the liquid's kinematic viscosity at the pumping temperature, and its specific gravity. Applied under ANSI/HI 9.6.7, these data produce corrected head, flow, efficiency, and power.

Manufacturers almost always publish centrifugal pump performance curves based on tests with water. If the actual liquid is more viscous than water — a heavy fuel oil, a syrup, a polymer solution — the pump will produce less head and flow, run at lower efficiency, and draw more shaft power than the water curve predicts. Correcting the curve is a standardized exercise, and it hinges on a modest but specific set of input data: a water performance curve from the manufacturer, best efficiency point (BEP) values read from that curve, the liquid’s kinematic viscosity at the pumping temperature, and its specific gravity, together with basic configuration facts such as stage count and operating speed. The governing reference is ANSI/HI 9.6.7, the Hydraulic Institute guideline for the effects of liquid viscosity on rotodynamic pump performance.

Why a water curve cannot be used directly for a viscous liquid

Pump performance curves describe the head added to the liquid, the shaft power, and the net positive suction head required (NPSHr) at a range of volumetric flow rates. Manufacturers obtain these values with water as the test fluid, even when the pump is intended for a different liquid, which is why thicker fluids require the curves to be adjusted (Pumps & Systems, June 2022). A viscous liquid increases friction in the impeller channels and running clearances and adds disc friction losses around the rotating impeller.

Two consequences are easy to miss. First, the performance penalty stems from friction, not density: pumping a fluid more viscous than water reduces performance even when the fluid’s reference density is the same as water, as the documentation for Applied Flow Technology’s Fathom software notes in describing the method. Second, the correction only applies to water-based curves. If the manufacturer supplies a curve for your working fluid at the required viscosity, that curve should be entered directly and any correction feature left disabled; otherwise the viscosity effect would be double counted.

The governing method: ANSI/HI 9.6.7

The correction procedure used across North American pump engineering is ANSI/HI 9.6.7, Rotodynamic Pumps — Guideline for Effects of Liquid Viscosity on Performance, published by the Hydraulic Institute. The edition referenced in current sources is ANSI/HI 9.6.7:2021, which carries forward an approach formalized in the 2010 edition, Guideline for Effects of Liquid Viscosity on Rotodynamic (Centrifugal and Vertical) Pump Performance. A rotodynamic pump is any pump that transfers energy to the liquid continuously through a rotating impeller — an umbrella that covers centrifugal, mixed-flow, and axial-flow designs; centrifugal and vertical pumps are the main subjects of this guideline.

The method is empirical. Correction factors were derived from test data on many pumps, and the standard defines ranges of validity — in viscosity, flow relative to BEP, and related pump parameters — outside which the correlations should not be extrapolated. It is written for Newtonian liquids, meaning fluids whose viscosity does not change with the shear they experience. Applied as directed, the method effectively transforms a water curve into a viscous curve, and commercial hydraulics software such as AFT Fathom implements it directly from the 2021 edition.

The data you need before applying the correction

InputWhat to collectWhy it is needed
Water performance curveHead versus flow and efficiency (or power) versus flow at the rated speed and impeller diameterProvides the baseline that the correction factors modify across the whole curve
Water BEP dataFlow at BEP, head at BEP, and efficiency at BEPAnchors the correction calculations; for multistage pumps, head per stage is used
Kinematic viscosityLiquid viscosity in centistokes or SSU at the pumping temperatureDrives the magnitude of the head, flow, and efficiency reductions
Specific gravityLiquid density relative to water at the pumping temperatureConverts corrected head, flow, and efficiency into shaft power
Pump configurationNumber of stages, operating speed, impeller trim, specific speedConfirms the pump falls within the guideline’s validity limits
Operating rangeThe flows and viscosities of interestDetermines which corrected curves or duty points to tabulate

1. A certified water performance curve

Collect the full curve rather than a single rated point: head versus flow, plus efficiency or brake power versus flow, at a stated speed and impeller diameter. The HI method positions each operating point relative to the pump’s BEP, so the shape of the curve across the working range matters. Confirm in writing that the curve is a water curve for the impeller trim you intend to run; a curve published for maximum diameter must be trimmed with the manufacturer’s guidance before any viscosity work. If the curve is given in normalized form (percent of BEP flow), keep that information available, because the calculations use flow positions normalized to BEP.

2. Best efficiency point data

The BEP is the flow rate at which the pump converts shaft power into useful head most efficiently — the peak of the efficiency curve on the water test. Three values from the water curve feed the correction: BEP flow, BEP head, and BEP efficiency. The correction calculations combine these with viscosity, so an imprecise BEP flow or head propagates error into every corrected point. For a multistage pump, head at BEP is evaluated per stage — total head at BEP divided by the number of stages — because each impeller produces its share of the total head.

3. Kinematic viscosity at the pumping temperature

The guideline works with kinematic viscosity (dynamic viscosity divided by density), the property that governs friction losses. It is reported in centistrokes, where one centistoke equals one square millimeter per second, and US-customary calculations conventionally convert it to Saybolt Seconds Universal (SSU), an older viscometer scale that underlies the familiar HI charts. Standard conversion tables handle the centistroke-to-SSU step; the relationship is not a single multiplier across the whole range, so use a validated conversion or the guidance published with the standard. Temperature matters just as much: the viscosity of liquids, especially oils, falls steeply as temperature rises, so enter the viscosity at the temperature of the liquid in the pump, not at storage or laboratory reference conditions. A viscosity-versus-temperature table from the fluid supplier is the most dependable source for this value.

4. Specific gravity

Specific gravity — the liquid’s density relative to water — is the input that turns corrected hydraulics into corrected power. In US units, the standard relation is that hydraulic horsepower equals flow in gallons per minute times head in feet times specific gravity, divided by 3,960; brake power then follows by dividing by the corrected efficiency. Because a viscous liquid depresses efficiency while the pump still moves the same fluid, corrected power can rise even where the liquid’s density barely differs from water — one reason fluids with water-like density still require the correction. The same specific gravity value is needed when sizing motors, couplings, and drives for viscous duty.

5. Configuration and operating data

Round out the data set with the number of stages, the operating speed the curve represents, and the impeller diameter or trim, all of which help confirm the pump sits inside the guideline’s validity ranges. Specific speed — an index built from speed, BEP flow, and BEP head that characterizes impeller geometry — is part of that check. If the pump will run at a speed different from the curve’s, scale the water curve with the pump affinity laws first and apply the viscosity correction to the scaled curve. Finally, decide the operating range and the viscosities to examine; corrections are usually run across several flows and, where the fluid composition varies, across a sweep of viscosities rather than a single value.

What the correction produces, and what it does not

Applied correctly, the method yields, for each viscosity of interest, a corrected flow, head, and efficiency at each point of the water curve, with shaft power recalculated from specific gravity and corrected efficiency. In practice, the corrected head-flow curve sits below the water curve, efficiency drops noticeably as viscosity climbs, and brake power rises for a given duty. The guideline addresses head, flow, and efficiency; it does not provide an NPSH correction, so published NPSHr values should be treated as water-based figures. Suction margins for viscous service should therefore be kept conservative, with manufacturer advice or testing sought for demanding applications.

Common data mistakes that distort the result

  • Correcting a curve that is already viscous-corrected. The method presumes a water baseline; applying it to a manufacturer’s viscous-fluid curve double counts the effect.
  • Using viscosity at the wrong temperature. A viscosity measured at a cool reference condition overstates the correction for hot service and understates it for cold service.
  • Substituting dynamic viscosity for kinematic viscosity. The method needs kinematic viscosity; with dynamic (absolute) viscosity data, convert using the liquid’s density.
  • Using total head instead of per-stage head for multistage pumps. This misplaces the BEP anchor used by the correction calculations.
  • Dropping specific gravity from the power step or sizing motors from uncorrected water power, which understates the absorbed power of viscous duty.
  • Extending the method beyond its limits — to flows far off the water curve, viscosities outside the stated range, or non-Newtonian and solids-laden fluids whose effective viscosity depends on shear and is not a single number. Such fluids need a different evaluation approach.

Practical takeaways

The checklist is short but specific. In writing, obtain the water performance curve at the correct speed and impeller trim along with its BEP flow, BEP head, and BEP efficiency, dividing BEP head by the number of stages for a multistage pump. From the fluid side, secure a viscosity-versus-temperature table and the specific gravity at the pumping temperature, converting to SSU where the calculation format calls for it. Run the ANSI/HI 9.6.7 correction — by hand or with software implementing the 2021 edition — across the flows and viscosities of concern, and confirm the pump lies within the standard’s validity limits. Recompute power with specific gravity and corrected efficiency for motor sizing, keep generous suction margin because NPSHr is not corrected, and where corrected performance looks marginal, ask the manufacturer whether a curve for the specific viscous fluid is available, since that data should be used in place of any correction.

FAQ

If the pumped fluid has the same density as water, is a viscosity correction still needed?

Yes. The performance reduction comes from added friction and disc losses caused by viscosity, not from density. A liquid more viscous than water lowers head, flow, and efficiency even when its reference density matches water.

At what temperature should the liquid viscosity be taken?

At the pumping temperature — the temperature of the liquid inside the pump. Liquid viscosity changes steeply with temperature, so a viscosity-versus-temperature table from the fluid supplier is the most reliable source for the correct value.

Does the ANSI/HI 9.6.7 correction adjust NPSHr?

No. The guideline corrects head, flow, and efficiency and recomputes power from specific gravity and corrected efficiency, but it does not provide an NPSH correction. Suction margins for viscous service should be kept conservative and confirmed with the manufacturer where necessary.

Can I apply the HI correction to a curve the manufacturer already supplied for the viscous fluid?

No. That would double count the viscosity effect. Enter the manufacturer's viscous-fluid curve directly and keep any viscosity correction feature disabled.

Is the method valid for slurries or shear-thinning fluids?

No. The guideline is developed for Newtonian liquids, whose viscosity is independent of shear rate. Fluids whose apparent viscosity changes with shear, or slurries carrying solids, behave differently in an impeller and require separate evaluation methods.

References

  1. Hydraulic Institute. ANSI/HI 9.6.7:2021, Rotodynamic Pumps — Guideline for Effects of Liquid Viscosity on Performance. Parsippany, NJ: Hydraulic Institute. https://www.pumps.org/wp-content/uploads/2022/01/9.6.7d.pdf
  2. Clark, Kyle. How to Make Viscosity Corrections for Centrifugal Pumps. Pumps & Systems, June 8, 2022. https://www.pumpsandsystems.com/article/how-make-viscosity-corrections-centrifugal-pumps/
  3. Applied Flow Technology. Viscosity Correction (ANSI/HI 9.6.7-2021 method). AFT Fathom documentation. https://docs.aft.com/fathom13/ViscosityCorrection.html
  4. ANSI/HI 9.6.7-2010, American National Standard (Guideline) for Effects of Liquid Viscosity on Rotodynamic (Centrifugal and Vertical) Pump Performance — ANSI webstore preview. https://webstore.ansi.org/preview-pages/HI/preview_ANSI+HI+9-6-7-2010.pdf

Related Terms

Leave a Reply

Your email address will not be published. Required fields are marked *