Coil geometry and pressure drop
A coil selection gives you rows and water flow; the next question is whether the thing fits and what it costs to push air and water through it. This sizes the face, works out the tube-side velocity and friction, and prices both pressure drops from the geometry.
| Quantity | Value | Note |
|---|---|---|
| Face area | 0.48 m² | 2.5 m/s face velocity |
| Face width × height | 0.53 × 0.90 m | 6 rows deep |
| Tubes per row | 19 | 50 mm pitch |
| Circuits | 6 | 0.167 L/s each |
| Tube inside diameter | 11.90 mm | 0.40 mm wall |
| Water velocity | 1.50 m/s | Reynolds 1.198e+4 |
| Friction factor | 0.0302 | Blasius, smooth tube |
| Water pressure drop | 17.5 kPa | Friction plus return bends |
| Air pressure drop | 119 Pa | 3.8 · rows · v^1.8 |
| Water volume | 2.1 L | Charge and freeze protection |
| Air-side area | 115.6 m² | 241 m² per m² of face |
| Tube-side area | 2.27 m² | All tubes, every row |
| Air film h | 70 W/m²K | j = 0.0102 at Re 1769 |
| Water film h | 5249 W/m²K | Dittus–Boelter, turbulent |
| Overall U | 37.6 W/m²K | UA 4.35 kW/K |
| NTU and bypass factor | 3.00 · 0.050 | Geometry drives the thermal result |
| Rated point | 24.0 kW | NTU 2.18, UA 3.16 kW/K |
| Half load | 0.028 | UA 2.60 kW/K at 1.25 m/s, water 8.5 °C |
Water side is Darcy–Weisbach; the air side is a calibrated correlation, so replace the coefficient with manufacturer data when you have it.
A = V̇ / v_face · Sets the coil footprint.
v = q̇ / (π d_i²/4) · Per circuit, not the whole coil.
f = 0.316 Re^−0.25 · Blasius, smooth tube, Re below 10⁵.
(f L/d_i + ΣK) ρv²/2 · Return bends carry a K of about 1.5 each.
A clear path from inputs to results.
A coil selection gives you rows and water flow; the next question is whether the thing fits and what it costs to push air and water through it. This sizes the face, works out the tube-side velocity and friction, and prices both pressure drops from the geometry rather than a rule of thumb.
How to use
- Set the airflow and the face velocity you want to run at.
- Enter the rows, face height and tube pitch to fix the tube count.
- Choose the tube diameter, wall and number of circuits.
- Enter the water flow and temperature.
- Read the face area, water velocity, and both pressure drops.
The formula
Reference: Darcy–Weisbach with the Blasius friction factor; air-side correlation calibrated to published finned-coil curves.
Worked example
1.2 m³/s coil, six rows, six circuits
- Airflow 1.2 m³/s at 2.5 m/s face velocity.
- Six rows, 0.9 m face height, 50 mm tube pitch, 12.7 mm tubes with 0.4 mm wall.
- Six circuits, 1.0 L/s of water at 7 °C.
- 1
Face area 1.2 / 2.5 = 0.48 m², so the coil is 0.53 m wide by 0.9 m high.
- 2
19 tubes per row at a 50 mm pitch; six circuits share the 1 L/s.
- 3
Each circuit carries 0.167 L/s through an 11.9 mm bore: 1.50 m/s and Re 11,980.
- 4
f = 0.316 × 11,980^−0.25 = 0.0302 over a 3.2 m developed length plus five return bends.
- 5
Water drop 17.5 kPa; air drop 3.8 × 6 × 2.5^1.8 = 119 Pa, about 0.22 kW of fan power.
0.48 m² face · 1.50 m/s water · 17.5 kPa water · 119 Pa air.
Interpreting the result
The water side is real hydraulics: doubling the flow roughly quadruples the drop, and adding circuits divides the flow per tube so the drop falls sharply. The air side is a correlation, so treat it as a screening number and replace the coefficient with manufacturer data when you have a real coil to select.
Assumptions & limitations
- One tube per row per circuit, joined by return bends with a loss coefficient of 1.5.
- Smooth tubes, Blasius friction below Re 10⁵, water properties at the stated temperature.
- Air-side drop is an engineering correlation of about 20 Pa per row at 2.5 m/s wet.
- A fouling resistance of 0.0002 m²K/W is included in the overall coefficient.
- A rated capacity, when given, replaces the correlation with the conductance implied by the manufacturer's point.
Frequently asked questions
What water velocity should I aim for?
Between about 0.6 and 2.5 m/s. Slower risks poor distribution and fouling; faster erodes the tubes and costs pump power for little gain in heat transfer.
Is the air-side number reliable?
It is calibrated to published curves and is good to roughly ±25%. Enter a rated capacity and the model inverts the rating to a conductance instead, which is exact for that point.
Why does adding circuits cut the pressure drop so much?
The flow per tube falls in proportion, and the drop goes with the square of velocity. Twelve circuits instead of six gives a quarter of the drop, at the cost of more complex headers.
What about the water volume?
It sets the charge, the freeze protection and how fast the coil responds to a valve movement. A large volume is slower to react.
Methodology last reviewed 2026-10-21. Results are informational and not financial or engineering advice — see our disclaimer.