Of all the inputs to an armor design, the predicted scour depth does the most work and carries the widest uncertainty.
Where it comes from
Computed to HEC-18, by adding long-term degradation, contraction scour, and local scour at each foundation. See contraction, local, and degradation scour and how to read a scour evaluation.
For sites without a formal evaluation — most bank protection, most private work — it is estimated, and that estimate should be recorded as an assumption rather than absorbed as a fact.
What it decides
1. Toe treatment
The most direct consequence. Two approaches, covered in toe protection:
Keyed: set the bottom row below the predicted depth, so the toe stays buried and supported after the design event. Requires knowing the depth, and requires excavating in the wet.
Aproned: extend units horizontally along the bed with enough material that as scour develops at the outer edge, they settle into the hole and continue protecting.
Either way the predicted depth is the number the treatment is dimensioned from. A toe designed for a depth that turns out to be optimistic is a toe that gets undermined.
2. Stack count
The published layout method: protection depth divided by bag height gives the number of stacks. See layouts, single, stacked and keyed and the calculator.
The bank covers how the anticipated scour depth affects the number and placement of units.
3. Plan extent
The one most often underestimated, and it follows from the depth.
A scour hole is not a vertical-sided pit. It is closer to an inverted cone with sides running out at roughly the angle of repose of the bed material. A deeper hole is therefore a wider hole, and in sand the horizontal extent for each vertical unit of depth is substantial.
So the protected footprint is set by the depth, not by the structure. At a pier, coverage has to span the footprint the hole would occupy, which is why HEC-23 expresses riprap extent at piers as a multiple of pier width. See bridge pier scour.
Getting the depth right and the extent wrong produces armor that is deep enough and too small, and it fails at its edges.
Living with the uncertainty
Scour prediction is not precise. Abutment scour in particular is acknowledged as less reliable than pier scour, and every method carries assumptions about bed material, alignment, and debris that the site may not honour.
Practical responses:
Margin where it is cheap. Extending the plan extent is usually cheaper than deepening a keyed toe. Where a choice exists, spend the margin on extent.
Choose the tolerant detail. A launching apron accommodates a deeper hole than predicted by settling into it. A rigid, precisely keyed toe does not. This is one place where a conforming system has a genuine advantage: it follows the bed down rather than bridging a void. See rock bags vs concrete revetment.
Design for incremental addition. Where units can be added at a boundary later, an under-estimate is a maintenance item rather than a reconstruction. See repair and unit replacement.
Establish the baseline and monitor. The first inspection sets the datum against which a deeper-than-predicted hole becomes visible. See post-installation inspection.
Ask what would change the answer. Sensitivity to bed material and to the discharge estimate tells you where the risk is.
The case where the prediction is not the problem
If the dominant component is long-term degradation, the bed is not going to reach a depth and stop. It is going to keep lowering, and armor sized to a predicted depth sits on a bed that will pass through it.
That is not a sizing problem and it is not solved by a deeper toe. It needs grade control, sediment management, or addressing the upstream cause. See contraction, local, and degradation scour and stormwater outfall erosion.
Establishing which component dominates before designing the layout is the check that prevents the most expensive mistake in this whole cluster.
What to record
- Predicted depth, by component, and the method
- Bed material assumed, and whether it was sampled
- Whether debris was considered
- Plan extent, and how it was derived from the depth
- Toe treatment and its dimension
- Sensitivity, if it was tested