Pipe Fittings Pressure Drop Calculator

A fitting disturbs the flow — turns it, splits it, screens it — and the energy that disturbance consumes is measured in velocity heads. One velocity head is v²/2g: at 4.44 ft/s it is 0.305 ft of water, at 8 ft/s it is 0.994 ft. A fitting’s K says how many of those heads it destroys: K = 0.9 for a standard 90° elbow, 1.8 for a tee branch, 10 for an open globe valve, 2.5 for a basket strainer. Total fitting head loss is ΣK × v²/2g — one multiplication for the whole board.

The velocity-head currency explains a pattern beginners find bizarre: the same elbow “costs more” on a big pipe than a small one. It doesn’t — K is dimensionless — but at the same GPM a bigger pipe runs slower, so its velocity head is smaller and every K on it is cheaper. Push 400 GPM through a 6-inch line (4.44 ft/s) and each elbow costs 0.12 psi; squeeze the same 400 GPM through a 3-inch line (18.0 ft/s) and each elbow costs 2.02 psi — seventeen times as much, purely because velocity head went from 0.305 ft to 5.03 ft.

This is also why the friction-factor slider on the board matters only for the equivalent-length readout, never for the psi: K-losses don’t care what f is. The slider exists because L/D = K/f mixes fitting data with pipe data — change the pipe’s roughness and the same fitting “equals” a different number of diameters. Keep the two currencies separate in your head and the whole subject stays clean.

The default mental model — “pipe loses the pressure, fittings are rounding errors” — fails exactly where most real runs live. Take the board’s default job: 100 ft of 6-inch Sch 40 carrying 400 GPM through six 90° elbows. Straight-pipe friction at f = 0.02: 0.52 psi. The six elbows at K = 5.4 total: 0.71 psi. Fittings own 58% of the loss while occupying perhaps 12 ft of physical space on a 112 ft run.

The equivalent-length view says the same thing in feet: six elbows × 45D = 270 diameters = 136.5 ft of hypothetical 6-inch pipe, against the 100 real feet. If the drawing had said “225 ft of pipe, no fittings,” the pressure budget would have been identical. Estimates that ignore fittings aren’t slightly wrong on routed industrial runs — they are roughly half wrong.

When does straight pipe win back the lead? Past about 50 diameters per fitting, pipe length dominates: a 500 ft straight run of 6-inch at 400 GPM burns 2.6 psi, and even a dozen elbows add only 1.4. The crossover is why plant mains with sparse fittings can lean on straight-pipe charts, while skid piping and mechanical rooms — dense with bends, tees and valves — must be counted fitting by fitting. The board above makes that split visible as a share bar: watch it flip as you toggle fittings on a short run.

A branched network is a set of parallel pressure statements: every path from the header to an outlet must land at the same endpoint pressure, and the flow distributes itself to make that true. If one branch is 30% less restrictive, it takes disproportionate flow — starving its siblings — which is the physics behind the shower that goes cold when the washer starts.

Design for balance rather than throttle for it. Each branch’s resistance is its pipe friction plus its fitting sum — and the branch tee itself charges K = 1.8 for every circuit that turns into it. Two standard moves equalize branches: enlarge the header one size (lowering velocity head makes every downstream tee cheaper), and size each branch by its own flow so that branch-to-outlet losses match. Where natural loss is unavoidably low, a balancing valve adds a deliberate, adjustable K — a dumb resistor placed on purpose.

A worked branch check: a 2-inch branch carrying 30 GPM through 50 ft of Sch 40 PVC and two elbows runs v = 2.86 ft/s, velocity head 0.127 ft, pipe loss 0.29 psi, fitting loss 0.22 psi (ΣK = 3.6 with the branch tee) — about 0.51 psi total. Its 1½-inch twin branch at 15 GPM: v = 2.00 ft/s, losses ≈ 0.15 psi. The 0.36 psi mismatch is small enough that the flows will stay within a few percent of design — that is what “balanced on paper” means. Run each branch through the board and the imbalance is arithmetic, not opinion.

The K-method assumes the fitting passes whatever the pipe asks of it. Choked flow is the regime where that assumption dies: lower the downstream pressure further and the flow rate does not rise. In liquid lines the mechanism is cavitation — as velocity peaks at the restriction, local static pressure falls toward the liquid’s vapor pressure, the stream flashes to vapor, and the void collapses as shock waves that sound like gravel in the line and pit metal within months.

Valves are the usual choke point because they concentrate the whole ΔP into one small throat. A globe valve already costs K = 10 wide open; throttled to 25% travel, its effective K is 30–50 times higher and the vena contracta inside it is where vapor forms first. The field signature of incipient choking is a hiss that quiets when downstream pressure rises and a valve that no longer increases flow as you open the upstream side. For water near ambient temperature this rarely matters on long runs — but across a pressure-reducing valve dropping 80 psi in one step, cavitation-choked operation is the default unless the valve is sized or staged for it.

The practical workflow: keep any single component’s share of the total drop below roughly 40–50% for water duty, stage large drops across two valves or a valve plus orifice plate, and treat a hissing valve as a design finding rather than a maintenance one. When the goal is simply more flow and the pipe is the limit, verify with the pipe pressure drop calculator; when the fittings are the limit, the board on this page shows exactly which line item to delete.

A pump suction manifold. 800 GPM into an 8-inch suction line: 20 ft of pipe, two 45° elbows, one gate valve, one basket strainer, one straight-run tee. Velocity 5.13 ft/s, velocity head 0.409 ft. Straight pipe: f × (L/D) × hv = 0.02 × 30.1 × 0.409 = 0.25 ft. Fittings: ΣK = 0.8 + 0.15 + 2.5 + 0.6 = 4.05, so 1.66 ft. Total 1.91 ft = 0.83 psi — and because 87% of that sits between the strainer and the pump flange, it comes straight off NPSH available. Suction-side fittings are priced differently from discharge fittings for exactly this reason.

A rooftop hydronic header. 120 GPM through 60 ft of 4-inch Sch 40 steel with four 90° elbows, two tee-branches and one butterfly valve (K ≈ 0.9 when fully open, comparable to an elbow). Velocity 3.03 ft/s, velocity head 0.143 ft. Straight pipe: 0.02 × 178.8 × 0.143 = 0.51 ft. Fittings: ΣK = 3.6 + 3.6 + 0.9 = 8.1, so 1.16 ft. Total 1.67 ft = 0.72 psi — with fittings at 69%. Size the expansion and pump heads knowing that number and the header behaves; guess it and the far coil runs cold every January.