Heat Exchanger Pressure Drop: Causes, Calculation and Reduction
Heat exchanger pressure drop is the irrecoverable loss of operating pressure as fluid moves from the inlet to the outlet of the exchanger. It comes from wall friction, direction changes, and entrance and exit effects, and your pump or fan has to pay for every pascal of it. You calculate it side by side (tube side and shell side, or hot channel and cold channel), compare it with the allowable limit for that fluid, and adjust passes, tube count, plate gap or flow rate until both thermal duty and hydraulic limits are met.
This guide walks through the definition, the main causes, the calculation routes for shell-and-tube, plate and hairpin units, a worked example you can repeat, and the levers that bring pressure loss down.
TL;DR
Pressure drop is the irrecoverable loss between inlet and outlet, and it excludes any change in static head. It sets your pumping or fan power.
Tube-side loss is friction in the straight lengths plus return, nozzle and entrance/exit losses. Shell-side loss depends on the bundle and baffles.
More flow, more plates or smaller gaps raise pressure drop. More passes or larger gaps lower it, though larger gaps also lower heat transfer.
Design to the allowable pressure drop for each side, then iterate. Heat transfer and pressure drop always trade against each other.
A spreadsheet-level calculation is enough for a first pass. Use CFD when geometry is unusual or maldistribution is a concern.
Table of Contents
What heat exchanger pressure drop actually is
The Heat Exchange Institute standard defines pressure loss as the irrecoverable loss in operating pressure as fluid travels between its inlet and outlet boundaries. For the tube side, the loss includes the inlet and outlet nozzles, the channels and the tubes. For the shell side, it includes the nozzles and the tube bundle. Static head changes are excluded (HEI Standard).
Two practical points follow from that definition.
Pressure drop is a boundary-to-boundary number. If a vendor quotes 0.5 bar, ask where the boundaries are. Nozzle losses can be a large share in compact units.
Elevation is separate. If your exchanger sits 6 m above the pump, the static head is a system calculation, not exchanger pressure drop.
“Pressure loss” and “pressure drop” mean the same thing in this context. You will see both in datasheets and standards.
Why pressure drop matters
Every pressure loss inside an exchanger has to be overcome somewhere else in the process, which makes pressure drop lost mechanical energy. In practice it shows up in four places:
Pump or fan size and running cost. Hydraulic power is volumetric flow times pressure drop, before pump efficiency.
Downstream pressure. A process that needs a minimum pressure after the exchanger cannot accept an unlimited loss.
Flow balance. In parallel branches, a high-drop exchanger starves its own branch.
Noise and erosion. High velocities used to chase heat transfer can also cause vibration and wear.
The central tradeoff is heat exchanged versus pressure drop. You can add surface area, but the pressure drop may become unacceptable for downstream processes (Cornell processdesign). A related IOSR study makes the same point: surfaces that give higher heat-transfer coefficients usually give higher pressure drop, while lower pressure drop reduces pumping cost (IOSR Journal).
What causes pressure drop in a heat exchanger
The losses fall into two families, and nearly every method sorts them this way.
Distributed losses come from wall friction along the flow length. They grow with length, velocity and surface roughness, and shrink with hydraulic diameter.
Localized losses come from events: sudden contraction at a tube inlet, expansion at a tube outlet, turns at return headers, nozzles, and fittings. A 2023 study of tube-side drop lists friction, expansion at the tube outlet, contraction at the tube inlet and reverse-flow effects, and reports shell-side drop rising with mass flow rate in the exchangers it tested (De Gruyter, 2023).
For a hairpin exchanger, the total combines distributed losses along the inner-tube and annulus lengths with localized losses at the fittings connecting the pipes and annuli (De Gruyter, Heat Exchangers: Design, Applications, and Performance).
The design factors with the biggest effect
Factor | Effect on pressure drop | Effect on heat transfer |
Higher flow rate | Rises sharply | Rises |
More plates (plate exchanger) | Generally rises | Rises with surface area |
More passes | Falls in plate units, per the Cornell notes | Depends on the layout |
Smaller plate gap | Rises | Rises |
Larger plate gap | Falls | Falls |
The plate-exchanger trends come from the Cornell processdesign notes, which also say typical industry gaps run from 1.3 to 1.5 mm (Cornell processdesign). Higher Reynolds number brings more flow resistance, and a 2026 Scientific Reports study concludes that heat-transfer enhancement can come with higher pressure drop, more pumping requirement and more entropy generation (Scientific Reports, 2026).
Fouling belongs on this list too. A deposit narrows the flow path and roughens the wall, so pressure drop creeps up over months. If you see it climbing at constant flow, read our heat exchanger fouling checklist.
How to calculate pressure drop
The method depends on the exchanger type, but the skeleton is the same: total drop is the sum of distributed and localized parts.
The basic building blocks
For a hairpin exchanger, the De Gruyter text writes the total as ΔP = ΔP_dist + ΔP_loc. It expresses the distributed term with a friction-factor relation, and the localized term as ΔP_loc = Kρv²/2, where K is a loss coefficient and ρv²/2 is the velocity head in pressure units (De Gruyter).
For straight passages the friction-factor term takes the familiar Darcy form:
ΔP_dist = f × (L/D) × ρv²/2
Here f is the Darcy friction factor, L the flow length, D the hydraulic diameter, ρ the density and v the velocity. Velocity enters squared, so doubling flow through the same passage roughly quadruples the drop when f stays about constant. That one fact explains most surprises when someone raises throughput on an existing unit. The friction factor itself comes from the Reynolds number and wall roughness. If you need correlations for it, our guide to forced convection in pipes covers the flow regimes you will meet.
Shell-and-tube exchangers
For a multi-pass shell-and-tube unit, NPTEL separates tube-side drop into straight-tube friction loss and return loss from direction changes, so ΔP_T = ΔP_t + ΔP_rt (NPTEL). The shell side is harder, because flow crosses the bundle, goes through baffle windows and leaks around baffles.
Two methods dominate shell-side work. The IIT Delhi design notes describe Kern as the simplest route and Bell-Delaware as the most widely accepted method (IIT Delhi). A sensible workflow:
Use Kern for screening and quick comparisons of layouts.
Switch to Bell-Delaware when the design is serious, because it accounts for bypass and leakage streams.
Cross-check against the vendor rating.
Plate exchangers
Plate units get their drop from the corrugated channel, the port and manifold, and the number of passes. Published correlations are geometry specific. Most manufacturers supply constants for their plate patterns, and a generic formula only gives you an order of magnitude. If you do not have the plate data, ask the supplier for the rated drop at your flow and temperatures rather than trusting a generic constant.
Finned and compact cores
For finned exchangers, core pressure loss includes entrance, flow acceleration, friction and exit losses. It is normally calculated from the fluid density at entry and exit and from the core flow geometry (Heat Exchanger Design Handbook). With gases, density changes through the core matter, so do not use a single average density without checking.
A worked example you can repeat
This is an illustrative case with assumed values, meant to show the arithmetic rather than describe a real unit.
Given: water at about 998 kg/m³ flows at 1 m/s through straight tubes with a 20 mm inner diameter and a 3 m length. Assume a Darcy friction factor of 0.03 and a return-loss coefficient K of 1.5 for the end of the pass.
Velocity head: ρv²/2 = 998 × 1² / 2 = 499 Pa.
Friction loss: 0.03 × (3 / 0.02) × 499 = 0.03 × 150 × 499 ≈ 2,246 Pa.
Return loss: 1.5 × 499 ≈ 749 Pa.
Total for this pass: about 2,995 Pa, roughly 3.0 kPa.
Now the power. Flow per tube is 1 m/s × (π/4 × 0.02²) ≈ 3.14 × 10⁻⁴ m³/s. Hydraulic power is flow times pressure drop: 3.14 × 10⁻⁴ × 2,995 ≈ 0.94 W per tube. For 100 parallel tubes that is about 94 W of hydraulic power before pump efficiency.
Now raise velocity to 2 m/s to improve heat transfer. The velocity head becomes 1,996 Pa and, with f unchanged, the drop climbs to about 12 kPa. Hydraulic power per tube rises to roughly 7.5 W, eight times higher, because flow doubles and drop quadruples. The heat-transfer gain from the higher velocity is far smaller than that. This is why the tradeoff is real and why designers target an allowable drop instead of the highest possible velocity.
Setting an allowable pressure drop for each side
There is no universal number. The IIT Delhi notes say a design must respect the specified shell-side and tube-side limits, and the hydraulic analysis computes the pressure drop along with the pumping power or fan work needed to maintain flow (IIT Delhi). In practice the allowable value comes from the system around the exchanger:
Available pump head. Subtract the losses in piping, valves and other equipment. What is left is the budget.
Downstream pressure requirement. The process may need a minimum outlet pressure.
Fluid value. A viscous or fouling fluid is often placed on the side where a higher velocity and a larger allowance are acceptable.
Economics. Extra pump size and energy cost are weighed against exchanger area.
Write the limit for each side into the specification before you size anything. Then both the thermal and hydraulic results are judged against the same targets. If you want to check the thermal side of that balance, our heat exchanger effectiveness and NTU guide pairs naturally with this one.
How to reduce heat exchanger pressure drop
Start with the cheapest changes and move toward the more expensive ones.
Lower the flow rate if the duty allows it. Because of the squared velocity term, small reductions pay back quickly.
Increase plate gap or number of passes. The Cornell notes list both as ways to cut pressure drop, and a larger plate gap also reduces heat transfer (Cornell processdesign).
Adjust tube passes or tubes per pass. When calculated drops exceed allowable limits, NPTEL says to adjust the number of tube passes or the number of tubes per pass (NPTEL). Fewer passes with more tubes per pass lowers velocity and drop, but also lowers the tube-side heat-transfer coefficient.
Review the nozzles and headers. Oversized nozzle velocities add to the boundary-to-boundary drop. Check them early.
Keep surfaces clean. Fouling control protects the pressure drop you designed.
Add area instead of velocity. A longer or larger exchanger runs at lower velocity and lower drop for the same duty, at a higher capital cost.
After each change, recheck the heat duty. A solution that meets the pressure limit but misses the thermal target is not a solution.
When a spreadsheet is not enough
Hand correlations assume uniform flow. They get less reliable with unusual headers, strong maldistribution between parallel channels, or compact geometries where nozzle effects dominate. In those cases a CFD study of the flow path shows where the loss really sits. If you are working at the component level, the same thinking applies to electronics cooling, and we explain how to start with the flow and pressure-drop budget in our heat sink CFD article.
Our perspective
We see the same pattern in many early designs: the thermal calculation gets done carefully, and the hydraulic check gets done at the end, if at all. Then the pump is too small or the exchanger is oversized. Our view is that pressure drop deserves to sit next to heat transfer from the first iteration. Conduction, convection and radiation models, fluid properties and hydraulic loss should live in one workflow, so changing a pass count or a tube diameter updates both answers at once. That is what we built the Jewlz Engineering Toolkit to support: physics-based thermal analysis, CFD and heat exchanger design optimization in a single environment with a built-in property database. We want students and small teams to be able to run professional-grade checks without a licensing barrier getting in the way. Joel
Run the numbers in the toolkit
If you want to test pass counts, flow rates and tube layouts without rebuilding a spreadsheet each time, try our heat exchanger optimization software. It sits inside the Jewlz Engineering Toolkit, which also covers thermal analysis and CFD and fluid flow simulation.
Basic tools are free to use, so you can try a case before deciding whether the paid features are worth it for your work.
Common mistakes to avoid
Mixing up static head and friction loss. Pressure loss excludes static head changes, so keep elevation in the system calculation.
Forgetting localized losses. Return bends, nozzles and fittings can add a large share, especially in short exchangers.
Using one friction factor for every regime. Laminar and turbulent flow need different treatment.
Judging only at design flow. Check the turndown and the maximum flow case as well.
Ignoring fluid property changes. A viscous liquid that cools down through the exchanger can have a very different viscosity at the outlet, and drop changes with it.
FAQ
What causes pressure drop in a heat exchanger?
Friction along the flow passages causes the distributed part, and events such as inlet contraction, outlet expansion, return turns and nozzles cause the localized part. Higher flow rate, smaller flow passages, longer paths and fouling all raise it.
Is heat exchanger pressure loss the same as pressure drop?
Yes, in this context the two terms describe the same quantity: the irrecoverable loss in operating pressure between inlet and outlet. The HEI definition excludes any change in static head, so elevation differences are handled separately.
How do I reduce pressure drop without ruining heat transfer?
Change the layout rather than only the flow. Adding passes or enlarging plate gaps lowers the drop, but a larger gap also lowers heat transfer, so recheck duty after each change. Adding surface area lets you run at lower velocity for the same heat load.
Which method should I use for shell-side pressure drop?
Use Kern for fast screening, since it is the simplest route. Use Bell-Delaware for a design you intend to build, because it is the most widely accepted method and accounts for bypass and leakage flows.
Sources
IOSR Journal: Investigation on Performance of Shell and Tube Heat Exchangers
De Gruyter: Experimental investigation of heat transfer characteristics, 2023
De Gruyter: Pressure drop, Heat Exchangers: Design, Applications, and Performance
Scientific Reports: Thermo-hydraulic and second-law analysis, 2026
IIT Delhi: Design Fundamentals of Shell-and-Tube Heat Exchangers
Recommended
Comments