Head Loss Calculator

The two formulas use different math and different roughness numbers. Darcy-Weisbach uses the Reynolds number and absolute roughness (ε). Hazen-Williams uses only the C factor and ignores temperature and viscosity.

Small gaps of 10 to 20 percent are normal. Big gaps usually mean one of these:

• Your C value and your ε value do not describe the same pipe condition.
• Your flow is laminar or transitional, where Hazen-Williams does not apply.
• Your fluid is not water.

For design work, trust Darcy-Weisbach. It is the physics-based method.

Always the inside diameter. Flow only happens inside the pipe, so wall thickness does not count.

This matters a lot for steel and copper pipe, where the nominal size is not the real inside size. For example, 4-inch schedule 40 steel pipe has an inside diameter near 4.026 inches, and 1-inch schedule 80 PVC is closer to 0.957 inches.

Check a pipe dimension table for your exact schedule. Using the wrong diameter is the most common mistake in head loss math, because friction loss changes with roughly 1/D⁵.

No. A negative total means gravity gives back more energy than friction takes away. This happens on downhill runs.

You typed a negative elevation change, and that drop is bigger than the friction plus minor losses. The fluid can flow on its own with no pump.

Look at the friction head loss card instead. That number is always positive and tells you how much energy the pipe itself eats.

Pick Custom from the fitting list and type your own K value. Most fitting makers publish K values or equivalent length data in their catalogs.

Some quick guides if you have nothing else:

• Standard threaded elbow: 0.9
• Sweeping or long-radius bend: 0.3 to 0.6
• Fully open valve: 0.05 to 0.2
• Partly closed valve: 2 to 20 or more
• Reducer or increaser: 0.3 to 1.0

K values are approximate. If fittings drive most of your loss, get real data from the maker.

You can, but only for low pressure drops. Pick Custom fluid and enter the air or gas density and viscosity at your working pressure and temperature.

The math here treats the fluid as incompressible. That is fine when the pressure drop is under about 10 percent of the inlet pressure. Above that, the gas expands as it moves, the velocity climbs, and the answer drifts low.

For long gas lines or steam, use a compressible flow method instead.

Between Re 2300 and 4000, flow flips back and forth between smooth and swirling. No formula predicts the friction factor well in that band.

Two identical pipes can give different losses on different days. The calculator still gives you a number, but treat it as a rough estimate.

If you can, redesign to stay clearly laminar or clearly turbulent. Changing the pipe size or the flow rate a little usually pushes you out of this band.

The table shows what happens to total head loss when you change diameter or flow rate by 10 or 25 percent, all else equal.

Use it this way:

• If loss is too high, look at the diameter +25% row. That is roughly one pipe size up.
• If your flow may grow later, check the flow +25% row to see if the pipe still works.
• Compare the percent change column to spot which input controls your system.

Diameter almost always wins. Going up one pipe size cuts loss far more than smoothing the pipe or removing a few fittings.

Velocity head is V²/2g. It is the energy the fluid carries because it is moving, written as a height of fluid.

Every fitting loss is a fraction of that number. The K value is that fraction. A K of 1.0 means the fitting throws away one full velocity head.

This is why fast flow hurts twice. Friction loss climbs with V², and every fitting loss climbs with V² too. Slowing the flow with a bigger pipe cuts both at once.

Not fully. Here is how each method treats it:

• Manning: Full support. It uses the real wetted area and hydraulic radius. This is the method built for partly full pipes.
• Darcy-Weisbach: Partial. It applies the hydraulic diameter, which is a fair estimate but not exact for open channels.
• Hazen-Williams: Ignored. That formula assumes a full bore pipe.

For sewers, storm drains, and culverts, use Manning with the partially full setting.

Oil is thick. Dynamic viscosity sits in the bottom of the Reynolds equation, so high viscosity pushes Re down fast.

SAE 30 oil is about 290 times thicker than water at room temperature. The same pipe and the same speed can be turbulent with water and laminar with oil.

In laminar flow, roughness stops mattering and f = 64/Re. Head loss then grows straight with velocity instead of with V². Always use Darcy-Weisbach for oil.

They give the same head loss. The calculator converts between them with L_eq = K·D / f.

Pick the one that fits your workflow:

• K-value: Cleaner for mixed pipe sizes. K does not change with pipe size or friction factor.
• Equivalent length: Handy when you want one total pipe length to hand to a pump chart or a spreadsheet.

Note that equivalent length depends on f, so it shifts if your flow rate or pipe roughness changes. K values stay put.

You edited the roughness, the C value, or Manning's n by hand. When you do that, the material drop-down switches to Custom so it does not overwrite your number.

Your edited value is still used in the math. To go back to the standard preset, just pick your material again from the list and all three coefficients refill.

A common target for water mains is 1 to 4 feet of loss per 100 feet of pipe, which is a gradient of 0.01 to 0.04.

The Hydraulic Gradient card shows your value directly. Use it to compare pipes of any length on equal footing.

• Below 0.01: pipe may be oversized and costly.
• 0.01 to 0.04: usual sweet spot.
• Above 0.05: expect high pump costs, noise, and wear.

Long transmission lines often run lower to cut pumping bills. Short branch lines can run higher.

Yes. Rust, scale, and slime build up over years, so a 30-year-old pipe is much rougher than a new one.

Rough guides for aging:

• New steel: ε about 0.046 mm, C about 120.
• Steel after 20 years: ε 0.5 to 1.5 mm, C 90 to 100.
• Old cast iron with heavy scale: ε 2 to 4 mm, C 60 to 80.

Plastic pipe barely changes. When you design a system meant to last, use aged values so the pump still works in year 20.

No. The extra segments feature only handles pipes in series, one after the other, carrying the same flow rate.

For parallel branches, each path carries a share of the flow, but all paths lose the same head. That needs an iterative network solve.

A workaround: run this calculator once for each branch, guessing the flow split. Adjust your guesses until every branch gives the same head loss.

Head loss is a height of fluid. Pressure drop is a force per area. They are linked by Δp = ρgh.

The bridge between them is density. Ten feet of water is about 4.33 psi. Ten feet of mercury is about 59 psi, because mercury is much heavier.

Head loss is handy for pump curves, which are drawn in feet or meters of head. Pressure drop is handy for gauges and pipe pressure ratings.

High velocity brings noise, pipe erosion, and water hammer risk. For water lines, keep it under about 8 ft/s (2.4 m/s), and under 4 ft/s on pump suction lines.

Fixes, best first:

• Go up one pipe size. Velocity drops with the square of the diameter.
• Cut the flow rate if the system allows.
• Split the flow into two parallel pipes.

Bumping a 2-inch pipe to a 3-inch pipe drops velocity by more than half at the same flow.

Yes, mostly through viscosity. Water at 40°F is about twice as thick as water at 180°F.

That changes the Reynolds number, which changes the friction factor. For turbulent water flow, the head loss shift is usually a few percent, so it is small but real.

For oils, temperature is huge. Viscosity can drop tenfold over a 50°F rise, which can flip laminar flow to turbulent. Always enter the true working temperature for oil systems.

Darcy-Weisbach with Colebrook-White is typically within 5 to 10 percent for straight pipe when your inputs are good.

The biggest sources of error are not the formula:

• Wrong inside diameter.
• Guessed roughness or K values.
• Ignoring pipe aging.
• Fittings placed close together, which interact.

Engineers usually add a 10 to 25 percent safety margin on total head before picking a pump.

It shows where your pipe sits among all possible pipe flows. The x-axis is the Reynolds number and the y-axis is the friction factor.

Read it like this:

• On the steep dashed line at the left: laminar flow, roughness does not matter.
• On a sloping part of a curve: transition zone, both Re and roughness matter.
• On a flat part of a curve: fully rough flow, only roughness matters and f stops changing with flow.

If you sit in the flat region, faster flow will not change f, so head loss grows cleanly with V².