British Waterways Develops Best Practice Hydraulic Methods for British Canals

British Waterways (BW) manages over 2000 miles of canals and river navigations throughout the UK. Its network of waterways is diverse in character in terms of channel characteristics, aquatic habitats and historic structures present (i.e. with 4763 bridges, 60 tunnels, 397 aqueducts, 549 locks and 89 reservoirs). BW controls the supply and distribution of water to the canals in order to maintain navigation, conserve habitats and for sale to commercial customers. The network also provides a valuable land drainage function with 30% of the catchment area of England & Wales draining into a waterway owned or managed by BW.

BW recently embarked on two comprehensive studies to improve the understanding of canal hydraulics. The primary aim of these studies was to determine the key hydraulic properties of canal channels and structures. The study approach utilised a combination of hydraulic trials and ISIS modelling to:

• Appraise the validity and quantify the uncertainty associated with the application of Manning’s normal flow equation to canals.

• Determine appropriate Manning’s roughness coefficients for varying canal bed and bank types.

• Determine an appropriate hydraulic representation of bridges, tunnels & aqueducts.

• Evaluate the impact of vegetation on the effective channel roughness.

Key study activities included:

• Identification of study pounds that would facilitate the investigation of the above aims.

• Completion of hydrographic surveys to establish channel and structural geometries as well as vegetation extents along study pounds.

• Implementation of extensive trials to measure the hydraulic response of each study pound to a range of flows.

• Quantification of bed seepage and gate leakage via indirect measurements.

• Undertake model trials with ISIS to determine hydraulic parameters and their associated uncertainties.

Two study pounds were identified, namely, the summit pounds of the Llangollen Canal and the Bridgwater to Taunton canal.

Trials and Modelling of the Llangollen Canal

The Llangollen Canal runs from Cheshire, meandering through Shropshire farmlands and crossing the border into Wales, through increasingly hilly countryside to terminate alongside the River Dee just above Llangollen.

The canal form varies along its length, with channel types of rock cut, concrete section and clay puddle lined. Numerous and varied in-channel structures are present such as tunnels, aqueducts and bridges. The canal bank types also show significant variation with unprotected, concrete, timber piled and sheet piled types all present.

For this study pound, the emphasis was to determine the validity of Manning’s uniform flow equation for a vegetated canal and to determine appropriate roughness coefficients and associated uncertainties for various bed and bank types. The study would include specifics to determine the hydraulic impact of aqueducts, tunnels & bridges.

Hydraulic trials were completed over a two week period utilising in-situ verified flow meters and remotely accessible automatic water level recorders. Flow was relatively simply controlled at the pound limits with sluices. There were only very limited sidelong feeder inflows present and these were measured throughout.

The Llangollen Canal was modelled along the 7.7mile study pound using the ISIS software. The model utilised 168 cross-sections in total with bridges resolved as Bernoulli losses, loss coefficients were derived from direct measurement. Tunnels and aqueducts were modelled as open channels with terminal Bernoulli losses. The model was calibrated between water level stations for a single flow rate. The model was then verified against eight other measured flow states.

The hydraulic trials and ISIS model application confirmed that the Manning’s normal flow equation is valid for un-vegetated canals and can be applied with a high degree of confidence. The study quantified friction coefficients for various channel types including;

• clay puddle lined (unprotected bank, sheet piled bank, vertical concrete wall, H piled with railway sleepers);

• concrete rectangular channel;

• annular brick lined tunnels with canter-levered towpaths;

• iron trough internal flange bolted aqueducts;

• quantified Bernoulli loss coefficients for tunnels, aqueducts and bridges.

Trials and Modelling of the Bridgwater to Taunton Canal

The Bridgwater & Taunton Canal was opened in 1827 and although only 14.5 miles in length, it passes through the characteristic lowland areas of Somerset, many parts of which have been chosen as Sites of Special Scientific Interest (SSSI) because of the variety of rare birds and plants which flourish there.

Similar trails to those conducted on the Llangollen Canal were completed for a 6.5 mile study pound along the Bridgwater & Taunton Canal. However, the emphasis here was with respect to the impact of varying vegetation extents. Consequently, trials were conducted for both winter and summer periods to quantify the seasonal variability.

The hydraulic trials were modelled using an ISIS model configuration similar to that for the Llangollen Canal, in total 161 cross sections were utilised. The model was employed to determining the Manning’s roughness coefficients for various trial flows. Roughness coefficients were determined in this way for both the winter and summer scenarios. The trails and modelling resulted in the development of generalised formulae to determine the hydraulic impacts of both marginal reed and pond weed.

Key Findings

From the above studies it was shown that the Manning’s normal flow equation may be confidently applied to evaluate the hydraulic response of canals throughout the UK. The study resulted in the quantification of friction coefficients for various channel types, this included unprotected and sheet piled banks, vertical concrete walling, timber piling, complete concrete rectangular channels, annular brick lined tunnels and iron trough aqueducts. Through the study, BW was able to determine the practical relationship to resolving the impact of margin reed on roughness and quantify head loss coefficients for tunnels, aqueducts and bridges.

BW is currently incorporating the results from these and other studies into a best practice guide for the hydraulic analysis of the UK canals. This will become available to BW’s engineers, consultants and other parties involved in the study of canals.