Design Considerations (Cont.)

Indirect Techniques for Erosion Protection
length from that point. Since design dike length would be shorter with this approach,
design spacing would be closer (see "spacing below"). The logic for this approach is
that the dikes must ultimately protect the newly deposited bankline. The weakness in the
logic is that if the dikes ultimately form a new bankline, then they will by definition, also
protect it. Therefore, the cost-effectiveness of this very conservative approach may be
questionable.
(c) Spacing and length are usually considered to be related, thus much of the literature
addresses the ratio of the two rather than separate values. In the absence of a need to
construct dikes to a predetermined channel alignment, the optimum length/spacing ratio
becomes a site-specific economic determination, involving a trade-off between shorter
dikes at a closer spacing against longer dikes at a greater spacing.
FHWA (1985) states that although spacing is a function of the length, angle, and
permeability of the dikes, as well as channel curvature, a parameter called "expansion
angle" may be used to better understand the relationship of these variables. In a straight
channel, for short dikes with permeabilities less than 35%, the expansion angle is the
same as for impermeable dikes, about 17 degrees. For permeabilities of 35% or greater,
the expansion angle increases as permeability or dike length increases.
FHWA (1985) also shows a method of determining dike spacing in a bend by
using a projection of a tangent to the thalweg at each dike tip. This procedure gives the
maximum allowable spacing, which should be decreased for a more conservative design,
particularly if short dikes or highly permeable dikes are used, if the banks are easily
erodible, or if the consequences of failure are high. They suggest that the expansion
angle be used to determine a prudent decrease in spacing from that which would be used
in a straight reach.
USACE (1981) and Copeland (1983) report a range in practice varying from a
spacing equal to dike length to a spacing of 6.3 times dike length, and describe USACE
model tests at the Waterways Experiment Station, Vicksburg, Mississippi, indicating that
the optimum spacing of impermeable dikes in a bend was between 2 and 3 times dike
length. However, they caution that those tests should not be applied verbatim to
practice, stating that "Spacing-to-length ratios for specific projects are best determined
by previous experience in similar circumstances or site-specific model studies." USACE
(1981) describes USACE model tests at the Missouri River Division's Mead Hydraulic
Laboratory, Nebraska, of very short impermeable dikes ("hard points") in a straight
channel, which indicated that flow downstream of each structure expanded at about a
20-degree angle from the main flow, a finding compatible with FHWA guidance. This
suggests that a spacing of about 3 times dike length for that type of dikes in a straight