“The spec calls for a given riprap D50. What bag is equivalent?”
It is the most common question on a substitution and there is no conversion table, for a reason worth understanding rather than working around.
Why there is no conversion
In riprap, the individual stone resists the flow. Its stability comes from its own submerged weight, its shape, and how it interlocks with its neighbours. Sizing methods relate hydraulic conditions to a required stone size, and D50 is the median of the specified gradation.
In a bagged system, the unit resists the flow. The mesh contains the fill, and the assembly behaves as one mass. The fill inside is not required to be individually stable, which is the whole point — it can be local rock rather than a quarried armor gradation. See when local stone changes the economics.
The two systems resist the same flow through different elements at different scales. There is no dimensional relationship between a median stone size in one and a unit size in the other, and anything presented as one would be a fabrication.
What to do instead
Go back one step, to the hydraulic condition.
The specified D50 was produced by a sizing calculation from a design condition — a velocity, or a shear stress, at a location, for a design event. That condition is the real requirement. The D50 is one system’s answer to it.
So:
- Find the hydraulic basis. What velocity or shear stress, at what location, for what event, produced this D50? It is in the design calculations
- Check what the number describes. Section-average or local? See depth-averaged vs local velocity
- Size from the condition, using the published rated conditions on product specifications and the calculator
- Document the route, so the substitution is auditable
That is a defensible chain. “The specification says this D50, so we propose this bag” is not.
When the hydraulic basis cannot be found
Common on older projects and on specifications inherited from elsewhere.
Options:
Ask. The engineer of record has it, or the design report does.
Re-derive it. Work back from the D50 through the sizing method that was presumably used, to recover the condition it implies. This is an inference and it should be labelled as one, because the method used may not be the one you assume.
Establish the condition independently. Get a design velocity for the site by the route in taking a design velocity from a model, and size from that.
Accept that the specification may not be substitutable. Sometimes the answer is that this is a riprap job. See when riprap is still the right choice.
The parts of the specification that do transfer
The D50 does not. Much of the rest does, and it is the part that decides whether the installation lasts:
- Plan extent — the area to be protected
- Toe treatment — keyed depth or apron dimension
- Terminations at flanks and crest
- Filter requirements, specified against the bed material
- The design event
Those are general armor requirements rather than riprap-specific ones, and a substitution should meet them all. See revetment anatomy and transitions and terminations.
Section thickness does not transfer directly. A riprap blanket thickness is a multiple of stone size; a bagged layout is stacks of units. The requirement it expresses — enough depth of protection — transfers; the dimension does not.
Making the substitution case
This is a procedural exercise as much as a technical one, covered in getting a non-standard product approved.
What tends to help:
- Lead with the hydraulic condition, showing you went back to the requirement rather than converting a number
- Address the transferable requirements explicitly — extent, toe, terminations, filter
- State the limitations. When not to use rock bags exists to be quoted, including by a reviewer testing us
- Be clear about what is not available: no measured test values, no certifications, no P.E. stamp, and no HEC-23 countermeasure chapter for this system. See HEC-18 and bagged armor design
The answer we will not give
A table mapping riprap gradations to bag sizes.
It would be the most requested page on this site and it would be an invention. The relationship does not exist, and publishing one would produce exactly the failure this site’s whole no-invented-numbers position exists to prevent: a plausible figure, on a page a reviewer relies on, pulled into a design without the conditions behind it.