Engineering & Technology · Water, Hydraulic & Coastal Infrastructure

Water supply, drainage and hydraulic networks

A pipe network must deliver the required flows at compatible pressures. A drainage network must receive and convey changing inflows under its downstream conditions. Both need conservation laws, but their boundary conditions and service goals differ.

Review fluid continuity, pressure and hydraulic head, energy per unit weight, groundwater/ground interaction, and area under a rate–time graph. Use SI units, with 1 L/s=.001 m³/s and 1 h=3600 s. Water density 1000 kg/m³ and g = 9.81 m/s² are declared where used. New channel, coastal and hydrological terms are introduced in the lessons.

Back to the chapter

After reading, you can

Apply mass and head accounts to networks while separating hydraulic service from water quality and drainage behaviour.

Conserve flow at nodes and energy along paths

For effectively incompressible steady flow at a junction with no storage, total incoming volume flow equals total outgoing flow. A tank adds a state: its inventory rises or falls when inflow and outflow differ. Demand and leakage belong in the node account; they do not vanish because a drawing has only connecting pipes.

Total mechanical head expresses mechanical energy per unit weight. It includes elevation, pressure and velocity terms. Friction and local disturbances dissipate head, while a pump can add head using external energy. A closed network needs flows that satisfy both node balances and compatible head changes around connected paths. Solving one equation for one pipe does not establish the whole network.

Pressure depends on the reference elevation

The hydraulic grade is elevation plus pressure head, HGL=z+p/(ρg). A point at a higher elevation has less pressure head at the same grade. Distinguish gauge pressure from absolute pressure and use a common vertical datum. A reported “40 m head” needs a statement of which head and reference are meant.

Head loss depends on flow, geometry and resistance. Its original value cannot simply be retained if demand, pipe condition or a connection changes. Steady calculations also omit transients: changes in momentum can create pressure changes that a sequence of steady states does not capture.

The figure supplies a path loss at a chosen flow instead of deriving it from a missing pipe diameter. This is useful for checking the pressure account. It is not enough evidence to select a pump, pipe, pressure limit or operating sequence.

Horizontal path distance 0–100 m, arbitrary original path, 5 units/m. Vertical elevation/head 0–60 m at 4 units/m. Supplied reservoir grade 50 m falls linearly to a low-velocity node grade 40 m, while node elevation is 20 m. Pressure head at the node is 20 m, or 196.2 kPa gauge using water density 1000 kg/m³, g9.81. Total lumped head loss10 m is supplied; node velocity head neglected. A different supplied loss18 m gives grade32 m and pressure head12 m, not plotted. No pipe diameter or operating condition is specified by this graphic.

Water quality and drainage add different claims

A drinking-water network carries constituents and water of different ages as well as flow. Hydraulic pressure and delivery volume alone do not establish quality. Source quality, treatment, storage, connections, integrity and maintenance all matter; a network model needs relevant observations and verification for its intended use.

Stormwater and wastewater systems may have free surfaces, backwater, surface flooding or pressurized sections. A submerged outlet can change the upstream response. Rainwater management may combine conveyance, infiltration, treatment and storage, with consequences for receiving waters. Keeping water types and interfaces explicit prevents a hydraulic connection from being mistaken for an acceptable service connection.

Models such as EPANET and SWMM represent particular network classes and processes. Their computed output remains conditional on topology, demands, boundary levels, resistance and time resolution. Calibration of flow does not automatically validate water quality, and a successful small event does not prove extreme-event performance.

Check one supplied path and a separate power account

Use a constructed water reservoir with surface grade 50 m above a common datum. Assume 10 m total head loss to a low-velocity node at elevation 20 m and neglect its velocity head. Node grade is 40 m; pressure head is 20 m. With ρ=1000 kg/m³ and g=9.81 m/s², gauge pressure is 196,200 Pa, or 196.2 kPa.

If a different supplied flow state instead has loss 18 m, grade becomes 32 m and pressure head 12 m, giving 117.72 kPa. The loss change is an input here, not derived from a complete network. At a separate no-storage junction, 20 L/s entering can balance 12 and 8 L/s leaving.

For a separate idealized pumping duty, Q=.020 m³/s through added head 15 m gives hydraulic power ρgQH=2943 W. At an assumed total electrical-to-hydraulic efficiency .70, input is about 4204 W. The calculation supplies neither an equipment rating nor evidence that the network meets all service conditions.

Try the reasoning

Does node flow balance alone determine pressure?
Show the answer and reasoning
No. The network also needs compatible head relationships and boundary conditions.
What happens to pressure at a higher node if hydraulic grade is unchanged?
Show the answer and reasoning
Pressure head decreases by the elevation increase, under the stated convention.
Does a matching hydraulic model establish drinking-water quality?
Show the answer and reasoning
No. Constituent behaviour, source/treatment conditions and relevant quality evidence are separate.

What to carry forward

Solve the appropriate conservation and head relationships, then check the service, quality and time-dependent conditions.

References

Teaching resources and research supporting this lesson’s explanations and examples.