Canal Irrigation & Design of Alluvial Canals

Canal systems

A canal is an artificial channel that carries water from a river or reservoir to the fields.

Classification

Canal alignment

Principles of alignment: command the maximum area with the shortest length; avoid villages, cemeteries, temples, valuable property; minimise cross-drainage works and balance cutting and filling; avoid rocky and unstable strata; keep curves gentle; place the head where the river is stable.

Canal losses

• Evaporation — small (a few per cent) except in hot, dry, windy conditions.
• Seepage — the major loss: absorption (wetting soil above the water table) and percolation (flow reaching the water table). Depends on soil permeability, water table depth, wetted perimeter, canal age (silting reduces loss) and velocity.
• Seepage raises the water table in the command and is a main cause of waterlogging — lining reduces it (see Canal Lining).

Canal cross-section

• Side slopes — in cutting roughly 1:1 (depending on soil); in filling flatter, about 1.5:1 to 2:1.
• Freeboard — vertical distance between full supply level and top of bank; increases with discharge.
• Banks — retain water, carry service road and inspection path; top width depends on discharge and road requirement.
• Berms — horizontal strips at the full supply level or the natural ground level between the bank toe and water; strengthen banks, protect against erosion, allow future widening. In partial cutting, the berm width is chosen so that the silt deposited on the side slopes forms the stable regime slope (about ½ : 1).
• Service road / inspection road on the bank; dowla — small bund on the road side for safety.
• Spoil banks — excess excavated earth dumped along the canal when cutting exceeds filling.
• Borrow pits — pits from which earth is taken when filling exceeds cutting; dug inside the canal bed (preferably) or at a safe distance outside the bank.
• Balancing depth — the depth of cutting at which the earth from cutting exactly equals that required for the banks (economical section).

Sediment and regime

Alluvial canals carry silt. If velocity is too low, silt deposits (silting); if too high, the bed and banks erode (scouring). A canal is in regime when it neither silts nor scours over a hydrological cycle.

• Initial regime — only the bed slope and depth adjust; width fixed by banks.
• Final (true) regime — all variables (width, depth, slope, shape) adjusted to the discharge and silt.
• Permanent regime — protected (lined) banks and bed; no further adjustment.

Kennedy's silt theory (1895)

R.G. Kennedy's observations on the Upper Bari Doab Canal (Punjab): silt is kept in suspension by the vertical eddies generated from the bed; so the silt-supporting power depends on the depth of flow (not width).

Design procedure (for given $Q$, $n$, $m$, $S$): assume depth $y$ → $V_0$ from Kennedy → $A=Q/V_0$ → find bed width (with side slopes ½ : 1) → compute $R$ and the velocity from Kutter's formula → compare with $V_0$; repeat until they agree.

Limitations:

• Considers only eddies from the bed (ignores those from the sides).
• No equation for bed slope — slope taken from Kutter's formula, which itself has limitations.
• B/D ratio chosen arbitrarily; trial procedure lengthy.
• Silt charge and grade are expressed only through the arbitrary ratio $m$.
• Based on data from one canal system.

Lacey's regime theory (1930)

Gerald Lacey found that regime channels have definite relations among discharge, velocity, hydraulic radius and slope. Silt is supported by eddies generated over the whole wetted perimeter.

Regime conditions assumed: constant discharge; constant silt charge; channel flowing in incoherent alluvium of the same grade that it transports (scoured and deposited with equal ease); unlimited incoherent alluvium.

Design procedure: find $f$ → $V$ → $R$ and $A$ → $P$ → solve for bed width and depth with ½ : 1 slopes → bed slope $S$.

Drawbacks of Lacey's theory: true regime is rarely attained; the concept of incoherent alluvium is idealised; the equations do not treat silt charge and grade separately; the single silt factor cannot represent both; semi-circular shape idea and empirical constants based on limited data.

Kennedy versus Lacey