Pipes in series and parallel, pipe networks

Head loss as rQ², rules for pipes in series and parallel, equivalent pipes, flow division, the three-reservoir problem and the Hardy Cross method for looped networks.

Drafted with Aria, reviewed by the AiCanCode.org team. Spotted an error? Use Give Feedback at the bottom of the page.

Why it matters

Real plants do not have single pipes. Cooling-water headers feed dozens of exchangers in parallel, utility lines change size along their length, and bypass lines run parallel to control valves. To predict how flow divides, or what a pump must deliver, you treat the system as pipes in series and parallel or, for looped networks, solve it iteratively with the Hardy Cross method.

Key ideas

Writing head loss in terms of flow rate. With V = 4Q/(πD²), Darcy–Weisbach becomes h_f = r·Q² with the pipe "resistance" r = 8fL/(π²gD⁵). At fixed f, head loss in turbulent flow varies as Q² and as 1/D⁵ — halving the diameter raises head loss 32-fold at the same flow. (In laminar flow h_f is linear in Q and varies as 1/D⁴.)

Pipes in series (end to end):

  • The same flow passes through every pipe: Q = Q₁ = Q₂ = …
  • Head losses add: h_total = h₁ + h₂ + … (plus any minor losses at the junctions, such as sudden expansions or contractions).
  • Equivalent pipe: a single pipe of chosen diameter D_e with the same total loss. With equal f, L_e = Σ L_i·(D_e/D_i)⁵ (Dupuit's equation).

Pipes in parallel (branching and rejoining):

  • The head loss across every branch is the same, because all branches share the same start and end junctions: h₁ = h₂ = … = h.
  • Flows add: Q = Q₁ + Q₂ + …
  • With h = r·Q², each Q_i = √(h/r_i), so Q = √h·Σ(1/√r_i). The flow divides in proportion to 1/√r_i: for equal f and L, Q ∝ D^2.5.
  • If one branch is closed with a fixed head across the system, the total flow falls and the other branches carry the same flow as before; if instead a pump must push a fixed total flow, the head across the system rises.
  • Note: the electrical rule 1/R = Σ 1/R_i applies only when the loss is linear in flow (laminar). For turbulent flow the rule is 1/√r_eq = Σ 1/√r_i.

Branching pipes / three-reservoir problem. Three tanks joined at a junction J: guess the head at J, compute each flow from its pipe equation, and adjust until continuity at J is satisfied (net flow zero). The direction of flow in the middle pipe depends on whether the junction head is above or below the middle tank.

Pipe networks and Hardy Cross. For looped networks two rules must hold:

  1. Continuity at every node: inflow = outflow.
  2. Energy around every closed loop: the algebraic sum of head losses is zero (clockwise positive). Hardy Cross: assume flows that satisfy continuity; for each loop compute the correction ΔQ = −Σ(r·Q·|Q|) / Σ(n·r·|Q|) with n = 2 for turbulent flow; apply ΔQ to all pipes of the loop (with sign) and repeat until corrections are negligible. Pipes shared by two loops get both corrections. Modern practice uses Newton–Raphson network solvers, but the method is standard in exams.

Hydraulic gradient. The slope of the hydraulic grade line, h_f/L, shows where the energy is being lost; in series pipes the HGL is steeper in the smaller pipe.

Formulas

  • h_f = f·(L/D)·V²/(2g) = r·Q², r = 8·f·L/(π²·g·D⁵) — h_f (m), Q (m³/s), r (s²/m⁵).
  • Series: Q₁ = Q₂ = Q, h = Σ h_i.
  • L_e = Σ L_i·(D_e/D_i)⁵ — equivalent length of a series combination at diameter D_e (equal f).
  • Parallel: h₁ = h₂ = h, Q = Σ Q_i, Q_i = √(h/r_i), 1/√r_eq = Σ 1/√r_i.
  • Q₁/Q₂ = (D₁/D₂)^2.5·√[(f₂·L₂)/(f₁·L₁)] — flow split between two parallel pipes.
  • ΔQ = −Σ(r·Q·|Q|)/Σ(2·r·|Q|) — Hardy Cross loop correction (turbulent).

Worked examples

Example 1 (standard). Water at 0.1 m³/s flows through 300 m of 0.3 m pipe followed by 200 m of 0.2 m pipe in series, f = 0.02 for both. Neglecting minor losses, find the total head loss and the equivalent length of 0.3 m pipe.

  1. r = 8fL/(π²gD⁵). r₁ = 8 × 0.02 × 300 / (9.8696 × 9.81 × 0.3⁵) = 204.0 s²/m⁵; r₂ = 8 × 0.02 × 200 / (9.8696 × 9.81 × 0.2⁵) = 1032.8 s²/m⁵.
  2. h₁ = 204.0 × 0.1² = 2.04 m; h₂ = 1032.8 × 0.01 = 10.33 m.
  3. h_total = 12.37 m.
  4. L_e = 300 + 200 × (0.3/0.2)⁵ = 300 + 1518.75 = 1818.75 m of 0.3 m pipe. h ≈ 12.4 m; equivalent to ≈ 1819 m of 0.3 m pipe. The short small pipe causes about 84 % of the loss.

Example 2 (GATE level). A total flow of 0.3 m³/s divides between two parallel pipes: A (500 m, 0.3 m, f = 0.02) and B (400 m, 0.2 m, f = 0.025). Find the flow in each and the common head loss.

  1. r_A = 8 × 0.02 × 500 / (π² × 9.81 × 0.3⁵) = 340.0 s²/m⁵; r_B = 8 × 0.025 × 400 / (π² × 9.81 × 0.2⁵) = 2582.1 s²/m⁵.
  2. Σ(1/√r_i) = 1/18.44 + 1/50.81 = 0.05423 + 0.01968 = 0.07391.
  3. √h = Q/Σ(1/√r_i) = 0.3/0.07391 = 4.059, so h = 16.48 m.
  4. Q_A = √(16.48/340.0) = 0.220 m³/s; Q_B = √(16.48/2582.1) = 0.0799 m³/s. Check: 0.220 + 0.080 = 0.300 ✓.
  5. Velocities: V_A = 3.11 m/s, V_B = 2.54 m/s — check that the assumed f values are consistent with Re and ε/D; if not, update f and repeat. Q_A ≈ 0.220 m³/s, Q_B ≈ 0.080 m³/s, h ≈ 16.5 m.

Common mistakes

  • Adding head losses for parallel pipes, or flows for series pipes.
  • Using D⁴ instead of D⁵ in the turbulent resistance (D⁴ is the laminar result at fixed Q).
  • Applying the electrical 1/R rule to turbulent parallel pipes.
  • Forgetting that a Hardy Cross correction applies with opposite sign to a pipe shared between loops.
  • Assuming f is the same in all pipes without checking Re and ε/D.

For GATE CH

Expect series and parallel numericals with given f (head loss, flow split, equivalent pipe), ratio questions (how flow divides between pipes of different diameters), the three-reservoir problem, and one Hardy Cross iteration on a simple loop. Practise writing h = rQ² quickly and solving for the split using 1/√r.

Quick check

  1. Two parallel pipes have equal L and f; D₂ = 2D₁. What is Q₂/Q₁?
  2. In series, if both pipes have the same Q, where is the HGL steeper?
  3. What is the energy rule around a closed loop?
  4. Halving D at fixed Q and f multiplies h_f by what factor?

Answers: 1. 2^2.5 = 5.66. 2. In the smaller-diameter pipe. 3. The algebraic sum of head losses around the loop is zero. 4. 32.

Interview questions

All Fluid Mechanics interview questions

Try answering each one aloud before you open it.

  1. 1.What is meant by pipes in series and pipes in parallel?Concept

    Pipes in series refer to a configuration where pipes are connected end-to-end, so the fluid flows through each pipe sequentially. In this setup, the flow rate remains constant throughout, but the pressure drop is the sum of the pressure drops across each pipe. Pipes in parallel, on the other hand, are arranged such that the fluid can split and flow through multiple paths simultaneously. In this configuration, the pressure drop across each path is the same, but the total flow rate is the sum of the flow rates through each pipe.

  2. 2.Explain how the equivalent resistance of a pipe network is calculated for pipes in series and parallel.Concept

    Write each pipe's loss as h = rQⁿ, with r = 8fL/(π²gD⁵) and n = 2 for turbulent flow. In series the same Q flows through every pipe, so resistances add: r_eq = Σr_i. In parallel every branch has the same h, so Q_i = √(h/r_i) and 1/√r_eq = Σ1/√r_i. The familiar electrical rule 1/R_eq = Σ1/R_i holds only when loss is linear in flow, as in laminar flow.

  3. 3.Why is it important to consider pipe networks in fluid mechanics?Application

    Pipe networks are crucial in fluid mechanics because they are used to transport fluids in various industrial applications, such as water supply systems, oil and gas pipelines, and chemical processing plants. Understanding how fluids behave in these networks helps engineers design efficient systems that minimize energy loss, ensure adequate flow rates, and maintain desired pressure levels. Proper design can lead to cost savings and improved system reliability.

  4. 4.What happens to the flow rate if one of the parallel pipes in a network is blocked?Application

    It depends on what fixes the operating point. If the head across the parallel set is fixed (for example, two reservoirs), the remaining branches carry exactly the same flow as before, and the total falls by the blocked branch's share. If a pump supplies the set, the system resistance rises, the operating point moves up the pump curve, total flow falls somewhat and head rises, so each remaining branch carries more than before. Either way the remaining branches see a higher share of the flow and the total is lower.

  5. 5.How does the diameter of a pipe affect the pressure drop in a series configuration?Application

    In series the same flow passes through each pipe, so velocity scales as 1/D². With Darcy–Weisbach at a fixed friction factor, h_f = 8fLQ²/(π²gD⁵), so head loss varies as 1/D⁵ in turbulent flow (1/D⁴ in laminar flow from Hagen–Poiseuille). Halving the diameter of one section raises its loss about 32 times, so the smallest pipe in a series line usually dominates the total pressure drop.

  6. 6.Explain the concept of hydraulic gradient in a pipe network.Concept

    The hydraulic gradient in a pipe network is a measure of the change in hydraulic head per unit length of the pipe. It represents the energy loss due to friction and other factors as the fluid flows through the pipe. The hydraulic gradient is important for determining the pressure distribution within the network and ensuring that the system can deliver the required flow rates at the desired pressures.

  7. 7.What is the impact of pipe roughness on fluid flow in a network?Application

    Pipe roughness affects the friction factor, which in turn influences the pressure drop in a pipe network. Rougher pipes have higher friction factors, leading to greater energy losses and higher pressure drops for a given flow rate. This can reduce the efficiency of the system and may require additional pumping power to maintain the desired flow rates. Engineers must consider pipe roughness when designing systems to ensure optimal performance.

  8. 8.Two pipes in laminar flow, with linear hydraulic resistances of 5 Pa·s/m³ and 10 Pa·s/m³, are connected in parallel. What is the equivalent resistance?Numerical

    In laminar flow Δp = R·Q is linear, so parallel resistances combine like electrical resistors: 1/R_eq = 1/5 + 1/10 = 0.3, giving R_eq = 3.33 Pa·s/m³. The flow splits in the ratio 2:1, with two-thirds through the lower-resistance pipe. For turbulent flow, where Δp ∝ Q², you would instead combine 1/√r values.

  9. 9.A fluid flows through two pipes in series with lengths of 10 m and 20 m, and diameters of 0.1 m and 0.2 m respectively. With the same friction factor in both, how do their pressure drops compare?Numerical

    At a fixed flow rate, Darcy–Weisbach gives h_f ∝ L/D⁵. For pipe 1, L/D⁵ = 10/0.1⁵ = 10⁶; for pipe 2, 20/0.2⁵ = 6.25 × 10⁴. So the 0.1 m pipe has 16 times the pressure drop of the 0.2 m pipe, even though it is half as long. In practice the smaller pipe also has a higher Re and possibly a slightly different f, which you would check.

  10. 10.Why might engineers choose to use a combination of series and parallel pipe configurations in a network?Application

    Engineers might use a combination of series and parallel pipe configurations to optimize the flow distribution and pressure management within a network. Series configurations can help manage pressure drops over long distances, while parallel configurations allow for flexibility in flow distribution and redundancy. This combination can enhance system reliability, allow for maintenance without shutting down the entire system, and improve overall efficiency by balancing flow rates and minimizing energy losses.

Finished this topic? Mark it so your progress, study plan and readiness keep up.

Stuck on something here?

PreviousMoody chart, pipe losses, fittings and equivalent lengthNext Boundary layer theory, drag and lift