§ 7.1  Module 7 — Matching and Optimising

The System Curve

A system curve is not a property of the pump. It is the bill the pipe presents to any pump connected to it.Module 7

7.1.1One intercept, then a rising bill

A pump is asked to overcome two different kinds of head. The first exists before any water moves: the difference between the suction and discharge water surfaces. The second exists only because water moves: straight-pipe friction, fittings, valves and the exit. Plot their sum against flow and the result is the system curve. Its governing statement is Hsys(Q) = Hstatic + Σhf(Q) + Σhm(Q).

At zero flow every velocity-dependent loss is zero, so the curve begins at static lift. That intercept is a useful audit: a spreadsheet whose first point is not the difference between water surfaces has mixed a pressure, elevation or datum into the loss schedule. Above the intercept, Darcy-Weisbach and minor losses both contain velocity squared. Because velocity is proportional to flow at fixed diameter, the curve rises approximately as Q². The friction factor changes slowly enough in fully turbulent municipal service that the parabola is a good picture, but a calculation should still recompute f at every flow.

Q = 0.080 m³/s
static lift                         18.00 m
straight-pipe friction               6.73 m
station and main fittings            1.27 m
                                    -------
system head at 80 L/s                26.00 m

quadratic teaching fit: Hsys = 18 + 1 250 Q²   (Q in m³/s)

A compact loss schedule for one condition; SI internally, flow converted once.

7.1.2Build it from the schedule, not from three guessed points

Choose a set of flows from zero through every credible operating flow. At each one, calculate velocity separately in every diameter, Reynolds number and Darcy friction factor for every straight segment, then each fitting loss at the velocity to which its K is referenced. Add the fixed static lift last. This is the same bookkeeping developed in Modules 3 and 4; the only new act is repeating it across flow. A curve assembled from a single design-point loss by square-law scaling is acceptable for an early comparison, but it conceals changes in friction factor and any element whose resistance changes with position.

A throttled valve does not add static lift. It adds a local loss and therefore steepens the curve above the same intercept. Aging does much the same through larger roughness. A lower wet-well level moves the entire curve upward by the added lift without changing the pipe-loss portion. These different motions matter in §7.2, because they move the operating point in different ways.

7.1.3Static-dominated and friction-dominated are behaviours, not labels

At the intended duty, compare static lift with the velocity-dependent losses. If 24 m of a 27 m duty is static, changing diameter can recover at most the remaining 3 m; a variable-speed drive cannot turn below the speed whose shutoff head clears the lift. If 6 m of a 27 m duty is static and 21 m is loss, diameter and speed have leverage, while roughness and valve position can move the duty dramatically. The test belongs at each important flow: the same main may be static-dominated at low flow and friction-dominated at peak flow.

7.1.4A curve is conditional evidence

The singular phrase “the system curve” is convenient and dangerous. A defensible design carries a family: high and low suction level, low and high receiving level, new and design-age roughness, valves in their normal positions, and every pump/header route that changes the loss schedule. A partially closed valve may be a commissioning test but is not a permanent energy strategy. An air pocket or blockage is not a legitimate operating curve at all; it is a fault to detect and remove.

Practice varies on which level combinations and pipe-aging assumptions a regulator will require. Name the authority, source elevations and material assumptions on the curve sheet. The physics of adding head is universal; the chosen envelope is a design decision that must remain visible.