Bank protection is needed where there is the risk of erosion of the bank and where this erosion would cause economic or environmental loss. If there is sufficient space available, it may be possible to reduce the need for bank protection by re-profiling the bank to a flatter slope to reduce velocities and encourage good vegetation growth. Even if bank protection is still required, it may be less severe if a flatter slope can be achieved, or may only be required below normal water level.

Based on "A T Pepper, C E Rickard (2009): 8 Works in the river channel. 36p. In: UK Environmental Agency (2009): Fluvial Design Guide (Contains public sector information licensed under the Open Government Licence v3.0.)"

Where it is needed, erosion protection for a bank can range from a good grass cover to heavy concrete slabs, but is broadly categorised as ‘hard’ or ‘soft’. Soft bank protection is generally considered to be vegetation of various types, whereas hard bank protection consists of concrete blockwork, riprap or similar. This is not a universally accepted definition as, in some parts of the country, riprap is termed ‘soft’ (because it is locally sourced and more natural than concrete), with ‘hard’ being reserved for the likes of piling and solid concrete walls.

Bank protection also has an impact on the ecology of a reach, and in this respect the softer the bank protection the more ecologically friendly it is likely to be. Vegetated banks offer little or no restriction on habitat, although animals that dig holes may need to be discouraged. This can be achieved by laying an open geotextile on the bank before adding a layer of topsoil (30–40mm thick) mixed with grass seed over the top. The geotextile assists in binding the grass roots together, at the same time making burrowing more difficult (though not impossible for determined creatures with sharp teeth).

Coir rolls, willow spiling and faggots can all be used to provide erosion protection, and each can also provide habitats for a range of species. However, the softer forms of erosion protection are not appropriate for situations where flow velocities or turbulence are high. Whereas the achievement of an environmentally acceptable protection system is very important, this has to be in the context of achieving the primary aim of bank stabilisation. A revetment that is destroyed in the first major flood because it was not up to the job would fail to achieve both structural and environmental objectives.

Some hard engineering solutions such as riprap can also provide habitats for a different range of species, both above and below the water, as the interstices provide shelter for fish fry and invertebrates below water, and for a range of insects, birds, animals and plants above water. Concrete slabs and solid concrete lining, on the other hand, provide a somewhat barren environment for flora and fauna.

The common approach to protecting the toe of a riverbank from erosion is to provide some form of flexible protection such as a gabion mattress or riprap. Alternatively, the toe can be protected by sheetpiling, with the top level of the piles below normal water level. Protection of the bank toe may be achieved in some cases by the use of in-channel structures such as vanes or groynes to deflect high velocities from the bank, and ideally cause a deposit of material in their lee to help protect the bank toe. These may be especially useful where the bank is well vegetated above the normal water level, and not, in itself, at risk of erosion. In floods these vanes or groynes are fully submerged, so offer minimal reduction in flow capacity.

Restoration techniques

The River Restoration Centre (RRC) collates information on river restoration schemes, and disseminates this information in a number of ways, one of which is the Manual of river restoration techniques (RRC, 2002). This should be among the first references to be consulted when considering a river restoration project.

Careful consideration is needed when proposing the removal of permanent artificial obstructions from a watercourse (such as a redundant weir) in order to restore a river to a previous state. Although there may be an immediate gain in flood conveyance and the removal of an obstruction to fish migration, there are also potential negative impacts. These include the release into the flow of accumulated sediments, which may be contaminated if the site has an industrial history. Removal of a weir may also be seen as a loss of amenity. It can also temporarily destabilise the river by causing bed erosion upstream, and this could undermine infrastructure such as bridges and riverside walls. But in many circumstances, a weir can be removed with no adverse – and many beneficial – impacts.

If a weir is to be removed, consideration must be given to the changes in the pattern of energy dissipation that may be caused. A weir pool may have formed over many years, with a shape and size that allow it to act as an effective energy dissipator. If removal of a weir takes away that function, that energy must be dissipated elsewhere – which is along the bed and banks of the river upstream. Careful assessment is required to determine the mean and maximum velocities likely to be encountered and the shear forces generated at the bed and banks. If these are likely to cause unacceptable erosion, then either the areas at risk should be protected or perhaps the weir should be only partially removed.

An alternative is to replace one large weir with a series of lower weirs or possibly a cascade, which would permit fish passage and improve flood capacity while restricting velocities and potential bed erosion. In less steep rivers, a pool and riffles sequence can be established, often by the introduction of imported gravel to form the riffles. Guidance on the establishment of pools and riffles can be found in the Environment Agency’s Guidebook of applied fluvial geomorphology (Sear et al, 2003).

In some situations, a meandering channel with pools and riffles can replace a shorter straight channel containing a weir, to the benefit of fish passage, an increase in habitat potential and additional flood attenuation. Where restoration involves reintroducing meanders to a long straight channel, it is likely that the new channel will cut across the existing channel in a number of places. If possible, parts of the straight channel that are intersected should be left open rather than being backfilled to provide backwaters. These will provide a still water habitat, as well as offer refuges for fish during periods of high flow. If land constraints permit, purpose-made backwaters or perhaps shallow bays can be excavated.

River diversions

Sometimes it is necessary to divert a river or stream course to make way for major infrastructure such as a motorway. This should be seen as the opportunity to re-create the river corridor, incorporating existing and new environmental features to ensure the new reach of channel adds value to the environmental and ecological status of the stream in question. The Water Framework Directive requires this type of approach. The river restoration techniques described above can be adopted for river diversion schemes (Fisher and Ramsbottom, 2001).

In other cases it may be that the existing course of the river is too constrained to allow enlargement to carry flood flows but is adequate for low flows. In such cases an alternative route for excess flows must be found. This may be a new channel, although this can involve significant land acquisition for something that is rarely used, or it can be a floodway where land is used for its original purpose (except when a major flood occurs) provided that such use is compatible with occasional flooding.

Literature sources
Fisher, K and Ramsbottom, D (2001). River diversions – a design guide. Thomas Telford.
River Restoration Centre (2002). Manual of river restoration techniques [online]. RRC. Available from:
Sear, D A, Newson, M D and Thorne, C R (2003). Guidebook of applied fluvial geomorphology, Defra/Environment Agency Flood and Coastal Defence R&D Programme, R&D Technical Report FD1914. Defra. Available from:
Measure category