Compute the local pressure losses that happen at every pipe fitting — bends, valves, tees, sudden contractions — with the K-factor method, in real time. Move the flow velocity, pipe diameter, ΣK and straight-pipe length sliders and see at a glance how the minor and major losses share the total, so you can size pumps and trim pipework with confidence.
Parameters
Flow velocity v
m/s
Pipe inner diameter D
mm
Sum of K factors ΣK
Add up the K of every fitting, valve, entry and exit
Fluid density ρ
kg/m³
Water ≈1000, gasoline ≈740, seawater ≈1025
Straight-pipe length L
m
Total length of straight pipe between fittings
Results
—
Velocity head v²/(2g) (m)
—
Minor head loss h_minor (m)
—
Friction head loss h_major (m)
—
Total head loss h_total (m)
—
Minor pressure drop (kPa)
—
Minor-loss share (%)
—
Pipe system view — pressure dropping in steps at each fitting
Fluid moving along the pipe loses pressure gradually along the straight runs (friction = major loss), and in sharp steps at every fitting (contraction, elbow, valve, tee = minor loss). The upper plot is the pressure profile along the pipe; the lower view shows the pipe and the flow particles.
Loss breakdown — major (friction) vs minor (local)
Each fitting's minor head loss equals the velocity head v²/(2g) times its dimensionless K factor. "Minor" historically means "local"; in short pipework these losses often exceed the straight-pipe friction loss.
$$h_{major}=f\,\frac{L}{D}\,\frac{v^{2}}{2g}$$
The major (straight-pipe friction) loss from the Darcy-Weisbach equation. f is the friction factor (this tool uses a representative f = 0.020), L is the straight-pipe length and D is the inner diameter.
The pressure drop from minor losses and the share of the total they carry. Short pipework crowded with fittings has a high minor-loss share.
What are pipe minor (local) losses?
🙋
When people say a pipe has "minor losses", I always assumed it just meant some small extra loss that doesn't really matter. Is that right?
🎓
That is the classic misunderstanding of the word. Pumping a fluid through a pipe network costs pressure for two distinct reasons. The first is the steady, distributed friction between the moving fluid and the pipe wall — it is spread along the full length of every straight pipe and is the "major loss". The second is the localised disturbance the fluid suffers every time it meets a fitting — a bend, an elbow, a tee, a sudden expansion or contraction, a valve, a pipe entry or exit, a flowmeter, a strainer. Each fitting yanks the flow into a new direction or area, creates turbulence and eddies that take energy from the mean flow, and so causes a small pressure drop concentrated at that one spot. These are the minor losses. The name comes from old textbooks, but "minor" really means "local", not "small".
🙋
Wait, so "minor" doesn't mean small? Then which one is actually larger most of the time?
🎓
Good question. In long-distance pipelines that run for kilometres with hardly any fittings, the major (friction) loss is indeed dominant. But in any plant piping system, HVAC duct, or pump suction header — short runs absolutely packed with valves, elbows and tees — the minor losses can easily make up more than half of the total and routinely exceed the friction loss. Try the length slider above and move L from 5 m to 200 m. You will see the share line completely flip between the two regimes.
🙋
So minor losses really aren't minor. How do you actually calculate them?
🎓
The workhorse is the K-factor method. Each fitting has a dimensionless loss coefficient K measured experimentally, and its head loss is h = K·v²/(2g). The v²/(2g) part is the velocity head — the kinetic energy of the flow expressed as a height. K values are tabulated in handbooks: a standard 90° elbow is about 0.75, a fully open globe valve 6 to 10, a fully open gate valve 0.15, a tee branch flow about 1.0, a sharp pipe entry 0.5, an exit 1.0. Add up the K's of every fitting in the system to get ΣK, multiply by v²/(2g), and you have the total minor head loss of the system. Simple, additive, and surprisingly accurate.
🙋
Nice — just add them up. What happens if you partially close a valve? Does K change?
🎓
It changes drastically with opening. A globe valve from fully open to half open can jump K by tens of times, and a ball valve at one-quarter closed can have K over 100. Every time you throttle a valve, the pump's electrical input is being burned as heat right there. That is why energy audits of plants always look first for valves that "live" half-closed — they are the cheapest energy savings around, usually fixed by switching the pump to a variable-speed drive instead.
🙋
What can a designer do to keep minor losses down in the first place?
🎓
Three big levers. (1) Route the pipework to use fewer bends, and replace sharp 90° elbows (K=0.75) with long-radius sweep elbows (K=0.45). (2) Pick valve types with low full-open K. Gate or ball valves (K=0.05-0.15) dump an order of magnitude less pressure than globe valves (K=6-10). (3) Avoid sudden expansions and contractions; use gradual tapered transitions instead, which can cut K by a factor of three to five. Try dropping ΣK from 10 to 3 on the simulator. At the same flow velocity, the pressure drop falls off a cliff and the pump duty drops with it.
Frequently Asked Questions
Pipe pressure losses split into two parts. The distributed friction along the pipe wall, proportional to length, is the MAJOR loss. The localised pressure drop at every fitting — bends, valves, tees, sudden contractions and expansions, entries and exits — is the MINOR loss. Each fitting is described by a dimensionless loss coefficient K, and its head loss is K·v²/(2g). In short pipe networks the so-called minor losses can dominate the total, so the name really means LOCAL, not small.
K factors are dimensionless numbers measured experimentally for each fitting shape and tabulated in handbooks. Typical values: 90° standard elbow ≈0.75, 90° long-radius elbow ≈0.45, tee through-flow ≈0.4, tee branch-flow ≈1.0, globe valve fully open ≈6-10, gate valve fully open ≈0.15, sharp-edged pipe entry ≈0.5, pipe exit ≈1.0. For a system you add all the K's to get ΣK, and the total minor head loss is ΣK·v²/(2g).
It depends strongly on pipe length. The major (friction) loss is f·(L/D)·v²/(2g) and scales with length, while the minor loss is ΣK·v²/(2g) and is independent of length. In short pipework crowded with fittings — plant piping, HVAC, building services — minor losses dominate; in long-distance pipelines with few fittings, friction dominates. Use the pipe-length slider in this tool to see the share flip between the two regimes.
Two conventions exist. The K-factor (velocity-head) method writes the loss directly as K·v²/(2g) and does not depend on the friction factor f. The equivalent-length method represents each fitting as Leq metres of straight pipe of the same diameter and folds it into the Darcy-Weisbach formula. They are related by K = f·Leq/D. The K-factor method is more general — especially for laminar flow, varying roughness, or non-Newtonian fluids — and is the recommended modern approach.
Real-World Applications
Process and chemical plant piping: Lines connecting reactors, heat exchangers, pumps and tanks are packed with flow-control valves, isolation valves, sampling tees, flowmeters and strainers. Even with only a few tens of metres of straight pipe, the system ΣK routinely reaches 30 to 80. The process engineer first estimates this minor loss with the K-factor method to size the pump's total head and shaft power. Miss it and a pump that runs at design speed simply will not deliver the rated flow — a textbook mistake that ends up costing time, money and pump replacement.
Building services — HVAC and water: Air-conditioning ducts and domestic water lines in a building branch out to every floor and room, piling up elbows, tees, balancing valves and check valves. For a typical water service, ΣK across the whole system can be 10 to 30, generating tens of kilopascals of pressure drop on its own. Underestimating it at the design stage means the top floor runs short of pressure and a booster pump has to be retrofitted later — exactly the kind of expensive rework K-factor sizing prevents.
Pump selection and energy audits: Even at the same flow and head, a centrifugal pump's efficiency swings sharply between shut-off, design point and overflow. Hang a pump on a piping system whose minor losses are far higher than estimated, and the operating point slides off the best-efficiency point and wastes electrical power. Energy audits always begin by re-examining ΣK, working out the real required head, and then trimming losses with variable-speed drives, low-K valves and gentler routing.
Sanity check before CFD: Before running OpenFOAM or ANSYS Fluent on a full pipe network, a K-factor hand calculation like this tool gives you a back-of-envelope estimate of where the pressure goes. If the CFD result is an order of magnitude different from the K-factor estimate, it is almost always a clue that the mesh quality, turbulence model or boundary conditions need a second look.
Common Misconceptions and Pitfalls
The biggest trap is "minor means minor, so we can ignore it". Skip minor losses in an early pipe-sizing calculation and, for short systems, you can miss most of the pump duty. With 10 m of straight pipe and ΣK = 20, leaving the minor losses out loses you more than half of the required head. The original English word "minor" means local, not small, and the system shown above with ΣK = 5 already has the minor loss carrying about 55% of the total at the default settings. Always include minor losses when the pipe is short and the fittings are many.
Next, "K factors are constant for any flow rate and diameter". The textbook K tables assume fully turbulent, fully developed flow (Re > 10⁴). In laminar flow, or immediately downstream of a bend where the velocity profile has not yet recovered, the effective K can be several times the tabulated value. Even the same "90° elbow" tends to a larger K at smaller diameters where relative roughness matters more. For critical designs, hunt down test data from the manufacturer at the Reynolds number you actually plan to run.
Finally, "add an equivalent length and use the Darcy-Weisbach formula for everything". The equivalent-length method Leq/D hides a hidden dependence on the friction factor f. Change fluid, pipe material or flow regime and the true Leq changes too, yet the tabulated value does not. In laminar flow, or with non-Newtonian fluids such as slurries and pulps, equivalent length can be wildly off. The K-factor method works directly with a physical loss coefficient and is the safer choice — and the one implemented in this tool.