AGGRADATION OF STREAMS DUE TO INCREASE IN SEDIMENT LOAD

Abstract

Aggradation occurs when the equilibrium of an alluvial stream is disturbed in such a manner that either the sediment carrying capacity of the stream is reduced or the rate of supply of sediment is increased over and above the carrying capacity of the stream. Aggradation is thus found to occur in many situations. The problem of aggradation due to supply of sediment, in excess of what the channel can carry, has been investi gated in the present study. The supply of additional sediment is assumed to be continuous and at a constant rate. The primary objective of the study is to provide a computational procedure for prediction of transient bed profiles on the basis of laboratory experiments. The analysis 0f experimental data provided also for the first time a clear understanding of resistance to flow and sediment transport in alluvial channels under non uniform flow condition. The time dependent variations of a river bed due to natural and/or human interference can be described by equations of motion for flow and equations of continuity for water and sediment. For large scale river morphological processes such as aggradation and degradation analytical models based on these equations have been presented by iii some investigators. The parabolic model proposed by de Vries, 9Z a2z viz.,, . _____ _ K .____ _ 0 5t 8x2 has been solved for the boundary conditions of the present problem and following expressions obtained : x Z = Z ( 1 - erf -_-—) . ... (i) 0 2YKt Z A-G - .'\P = 0.885 — ... (ii) fit K(1-A) 1 = 3.66 TfKt ... (iii) in which Z is the aggradation depth at time t at any distance x from the section of sediment injection; Z i o is the maximum depth of deposition at x = 0;AG is the sediment load at x = 0 in excess of the equilibrium sediment transport rate; K is the aggradation coefficient; 1 is the length of aggradation; and >, is the porosity of the sand mass. A tilting recirculatory flume of rectangular cross-section 20 cm wide and 30 -p long was used for experimental investigation of the problem. The sediment forming the bed and the injected material was natural sand with a median diameter of 0.32mm and a geom'etric standard deviation of 1 *30. After the establishment of uniform flow iv for a given discharge and slope, the sediment supply rate at the upstream end of the flume was increased to a known value by continuously feeding excess sediment at the upstream end of the flume. The bed and water surface profiles downstream of the section of sediment injection were recorded at intervals. The added sediment load was varied from 0.30 G to 4.0 G , where G^ is the equilibrium e c e sediment transport rate. The theoretical expression Z - z0(1~ erf x/."/^) i.e. Eq.(i) has been arrived at after many simplifying assumptions and it is to be expected that results from this shall not fit the experimental data directly. On comparison with the experimental data it has been found that the form of the equation is correct, but the value of the aggradation coefficient,K, enabling fit of this equation to the experimental data Js different from the 1 * c> theoretical value, KQ •^ g H,M ( Here b is ±YlQ 0 b exponent in sediment transport law of the form G = a U and S is the bed slope). This modified value of K (enabling fit of Eq(i) to the experimental data) has been empirically related to the theoretical value, KQ, and to the relative rate of overloading AG/GQ. As such this relation alongwith eq.(i) enables prediction of transient bed profiles, when the value of Zq has been computed from Eq.(ii) with the modified value of K. The extent of stggradation can be known from the Eq.(iii) using again the modified value of K. The analysis of non-uniform flow data obtained from aggradation runs has revealed the following facts: (i) The sediment transport lav/ valid for uniform flow conditions cannot be applied directly to non-uniform flow conditions obtained i'n aggrading streams, (ii) The i concept of lag of sediment transport proposed by Kennedy is only partially supported by the present data. (iii) Resistance lav/ under non-uniform flow, conditions is seen to be the same as for uniform flow provided the local friction slope is used in place of Sq in the former case. •