You are on page 1of 19

Volume 4 Section 2 Part 1 HA 71/95

Chapter 6 Assessment of Bridge Afflux

6. ASSESSMENT OF BRIDGE AFFLUX
Review of currently available methods for estimating afflux 6.1 There are six established hand methods of estimating bridge afflux, and at least three numerical methods which involve the construction of a computational model. 6.2 Hand methods other than the six listed below have not been widely published or applied: (a) “The computation of backwater and discharge at width constrictions of heavily vegetated flood plains”, US Geological Survey, February 1977. (Ref No.9) US Bureau of Public Roads (USBPR) Method, revised 1978. (Ref No.10) US Geological Survey (USGS) Method, originally developed between 1955 and 1958 and revised in 1967. (Refs No.11,12 & 13) The method published in “Open-channel hydraulics” by French in 1985. (Ref No.14) Yarnell’s method first published in 1934. (Ref No.15) Method for calculating afflux at arch bridges developed at HR Wallingford and published in 1988. (Ref No.5) fairly well developed stage. Available computational methods include: (a) A mathematical model produced by the Hydrologic Engineering Center (US Army Corps of Engineers, 1979). It is based on Yarnell’s experimental study. (b) Finite Element Model for Bridge Backwater Computation used in the evaluation of flood risk factors in the design of Highway stream crossings in the United States. It is a steady state model, developed to model wide valleys with heavily vegetated flood plains. (c) Models produced in the UK incorporating the USBPR method for calculating bridge afflux. These can model steady or unsteady flow and are applicable to blockage of flood plains as well as the main river channel. (d) Other computational models available in the UK allow head losses to be included to represent the effects of bridge constrictions. 6.5 When considering the use of computational packages to model flows in the vicinity of bridges, it is essential to understand the method used for afflux calculation. A computational model is only as good as the hydraulic formulae it implements. If inappropriate formulae are used the results will be incorrect, however impressive the presentation of the results. Recommended method for calculating bridge afflux 6.6 The general procedure includes the estimation of the design flood peak discharge, the estimation of the resulting water level on the existing river, followed by the estimation of afflux resulting from the proposed road design. 6.7 The calculation of design discharge is based on a simplification of the method contained in the Flood Studies Report, NERC (1975, 1985) (Ref Nos 3 & 16) and is described in Chapter 4. The calculation of the existing water level is based on the typical section method and is described in Chapter 5.

(b)

(c)

(d)

(e)

(f)

6.3 Method (b) is considered to be the most appropriate because it has been verified against a large number of floods and it is comparatively easy to use. It is of limited use for rivers with heavily vegetated flood plains, for example dense woodland, but these cases are comparatively rare in the UK. 6.4 None of the numerical methods are particularly quick or easy to use, although those already familiar with a particular program or package may find it easier to apply than to use the recommended hand calculation. The use of computational models as discussed in Chapter 8, is often appropriate when the design proposals are at

August 1998

6/1

the method lends itself well to the use of a basic spreadsheet. 2 and 4 respectively. so that the conveyance values calculated for the single section as described in paragraph 5. Some surveyed section data of the river channel and flood plain in the vicinity of the bridge must be obtained. These sections are shown in Figure H1. computed from the stream at Section 1 on Figure H1. 6. then it will be necessary to carry out a minor survey.17 The velocity distribution in a river varies from relatively high velocities in the deeper portions of the river channel to lower velocities near banks and on the flood plains. culverts will generally be required unless the flood plain is spanned by a viaduct.29 to 6. Qb and Qc. The same data are required as in the calculation of flood water level described in Chapter 5. The procedure for calculating afflux is summarised in Figure 6. A2 and A4 = Cross sectional area at Sections 1. It will be seen that the flow through the culverts is dependent on the value of afflux calculated. This type of calculation could be carried out with a hand calculator. In the example given in Figure 5. These data may be available from the River Authority but if not. the second term of the formula is negligible so that it is reasonable to use the approximation: The calculation of κ and α2 is described below.16 It is suggested in Figure 6. However. The flow through the culverts can be calculated in accordance with 6. the estimation of afflux will be based on a simplified version of the USBPR method. since it requires the building of tables of various quantities. as this is often the largest factor causing afflux.10 UK rivers are generally relatively uniform in cross section and the afflux for modern highway bridges is usually small compared to the total depth of flow.2. Under these conditions. A worked example is given in Appendix I.8 As discussed in paragraph 6. 6. then M can be calculated as: Qb/(Qa + Qb + Qc .15 If significant flood plain flow already occurs. The average velocity head.15 are used for the calculation of Qa.Chapter 6 Assessment of Bridge Afflux 6. where κ = overall backwater coefficient = velocity in the bridge opening V2 g = acceleration due to gravity α1 and α2 are the kinetic energy correction factors at Sections 1 and 2 respectively A1.11 This is the ratio between the existing flow which passes between the proposed locations of the new bridge abutments and the flow which passes between the abutments when the proposed bridge 6/2 August 1998 . 6. 6. Bridge opening ratio (M).14 It can be assumed that the river and flood plain cross section is constant throughout the length of the reach shown on Figure H1. in which there are no flood plain culverts.1 that the first iteration is carried out after the base curve coefficient has been calculated.1. M can be calculated as: M = Qb/(Qa + Qb + Qc) If culverts are present.9 The full equation for afflux (h1) is Volume 4 Section 2 Part 1 HA 71/95 is in place (ie the total flow minus flow through any proposed flood plain culverts). Kinetic energy correction factor 6.32 below.13 If the crossing is skewed in relation to the estimated general direction of flow. then Qa. The flow through the culverts is then subtracted from (Qa + Qb + Qc) which gives rise to a new value of M.12 In the case illustrated in Figure H1.3 above. 6. Therefore an iterative afflux calculation procedure is required if flood plain culverts are present. Qa = K1 s½ Qb = (K2 + K3 + K4) s½ Qc = K5 s½ 6. Qb and Qc are calculated based on the projected width of bridge opening. does not give a true measure of the kinetic energy of the flow. as shown in Figure H6 for example. 6.flow through culverts) 6.

∆κ for M = 1. the ratio of the area of pier blockage to the gross projected area of flow between the abutments. Graph (B) is then used to obtain σ.22 The pier blockage incremental coefficient (∆κp) is calculated in accordance with Figure H4. 6. α2 is the value at Section 2. it will be necessary to recalculate α1 when the culvert flow is deducted from (Qa + Qb + Qc).4 and therefore off the scale of Figure H2.19 It should be noted that if α1 is greater than 3. Skew incremental coefficient 6. is required to correct the velocity head for nonuniform velocity distribution under the bridge. above.21 The top curve shown is for abutments with 90 degree wingwalls. if the span of the bridge opening is greater than 60 metres. The value of α1 is calculated from the flow ∆κp = ∆κ σ areas and conveyance of each subsection which have already been calculated as in paragraph 5. α 2 is read off Figure H2. A second coefficient. 6.15. Pier blockage incremental coefficient 6. then it will be necessary to reconsider the width of the floodplain which is actually conveying flow. or Figure H6 for skewed crossings. is calculated as shown in the top part of Figure H4. 6. then the abutment/ wingwall type is not significant and the lower curve should be used for all types of abutment.38.20 The base curve coefficient κb is read off Figure H3. The value of ∆κs is dependent on the value of skew angle and on the value of M. In the calculation of An2 the normal water surface level is the pre-bridge construction level. ∆κp is then calculated as: where Q V A qi vi ai = total flow at section = average velocity at the section (=Q/A) = Total area of section = flow in a subsection i = average velocity in a subsection i (=qi/ai) = area of subsection i α1 is the value at section 1 as shown in Figures H1 and H6. α1 and α2 are related as shown in Figure H2.0. Base curve coefficient 6. The kinetic energy correction factor (á) at a section is calculated as: Chapter 6 Assessment of Bridge Afflux between 45 and 60 degrees. κb is dependent on the value of M and type of bridge abutment. This is discussed further in paragraph 6. Eccentricity incremental coefficient If flood plain culverts are present. 6.18 Using α 1 and the value of M calculated above. In addition. J.0 is then read off graph (A) on Figure H4 using the calculated value of J and the pier shape which corresponds most closely to the proposed pier design. at the bridge constriction itself. á2. by a kinetic energy coefficient. The value of e is calculated as follows: August 1998 6/3 .23 The eccentricity incremental coefficient (∆κe) is read off Figure H5. the correction factor for cases where M is not equal to 1.Volume 4 Section 2 Part 1 HA 71/95 A weighted average value of the kinetic energy is obtained by multiplying the average velocity head. the middle curve is for abutments with 30 degree wingwalls and the lower curve is for abutments with wingwalls angled Qa and Qc are illustrated in Figure H5. a1. The value of ∆κe is dependent on the value of eccentricity e and the value of M. If the abutment is without wingwalls (ie spill through type) then the lower curve should be used.24 The skew incremental coefficient ∆κs is read off Figure H7.

28 are repeated. 6.54 B g f Hu1. ie gross flow area between the abutments.5 κ = κb + ∆κp + ∆κ + ∆κs Velocity through bridge opening 6.31.25 Two cases are illustrated in Figure H7. The value of Hu is calculated from Hd plus the value of afflux calculated in 6.8Ac [2g(Hu .32 This approximate formula may be applied to any standard culvert shape.12 to 6. where Q is total flow minus flow through any flood plain culverts.31 The flow through a culvert which is submerged is calculated as follows: Q = 0. The flow through the culverts is subtracted from the total flow (Qa + Qb + Qc) and the value of M is recalculated as in paragraphs 6. 6.x L where h1 is the afflux at the crossing. Hence the afflux at x metres upstream of crossing: hx = h1. A2 is the flow area at the bridge constriction.5 where and 6/4 August 1998 . ie the value of afflux calculated in paragraph 6. eccentricity and skew. Backwater length 6.29 The flow through a non-submerged flood plain culvert is calculated as follows: Q = 0.28 is approximately equal to the value calculated in the previous iteration. Afflux 6. ie the flow which bypasses the main bridge opening. The flow through each culvert is calculated separately and then summed to give the total flow through the culverts.15. Diagram (A) should be used where the bridge abutments are aligned with the river flow and Diagram (B) should be used where the abutments are at 90 degrees to the axis of the structure.26 The overall backwater coefficient ê is the sum of the base curve coefficient and the incremental coefficients for pier blockage.33 Using the new value of M. Overall backwater coefficient 6.16 above.35 The afflux decays approximately in proportion to the distance from the proposed crossing. This iterative process is continued until a convergence to a value of afflux is found which approximately satisfies the conditions.28 above. steps 6.27 The velocity through the bridge opening (V2) is calculated as follows: V2 = Q A2 where Ac = Area of culvert 6. 6.Chapter 6 Assessment of bridge afflux B Hu Hd = = = Volume 4 Section 2 Part 1 HA 71/95 average flow width in culvert height of u/s water level above culvert invert height of d/s water level above culvert invert 6.30 The value of Hd is calculated from the value of existing water level at the road crossing location as calculated in paragraph 5. It should be possible to find an approximate solution in 2 or 3 iterations.18 to 6.Hd)]0.28 The bridge afflux (h1) is calculated as follows: h1 = κ α2 V22 2g Flow through flood plain culverts 6.34 The length of river affected by the afflux caused by the crossing may be estimated using the backwater length formula given in paragraph 5. 6.

There is a danger that the flow velocities in wide flood plain sub-sections are calculated as being very small which results in a very high value of α1. and thereby a large value of afflux h1. a severe reduction could not easily be justified.15 and then recalculate α1 in accordance with paragraph 6. Also it should be noted that only very limited tests were carried out for the situation where the bridge is located within the backwater influence of some constriction downstream. In practice the water levels will vary. if the value of α1 is greater than 3. particularly for skew August 1998 6/5 . including old arch bridges. the afflux produced by the combination of bridges will be more than the value for an individual bridge but less than the value which would result by considering the two bridges separately. It is recommended therefore that. For example. However. The value of α1 is dependent on the choice of subsections.42 The application of the method to “nonstandard” road layouts is considered with reference to the layouts shown on Figure 6.3m) are least likely to be conveying flow in the real river and could be omitted from the calculation of flow area. However. etc. but an initial estimate may be obtained from an analysis of adjacent bridges carried out by USBPR (1978) (Ref No. It should be noted that the method assumes constant water levels along the upstream and downstream faces of the road embankments. the maximum value of skew angle for which a value of skew incremental coefficient is Chapter 6 Assessment of Bridge Afflux given in Figure H7 is 45 degrees. In addition the laboratory and field tests were mostly carried out in situations where the river is relatively straight so that it is difficult to assess what the effect of a road crossing at or near a severe bend in the river would be.41 Adjacent bridges often occur where a new road scheme is constructed in the vicinity of existing bridges.40 Where two bridges are located close to each other. If two adjacent identical bridges over a river were moved closer and closer together the afflux would tend towards the value for a single bridge. Case (a) is the standard application for which the method has been designed.Volume 4 Section 2 Part 1 HA 71/95 Application and limitations of the method 6. assuming Figure H2 is extrapolated. 6. Examples of road layouts 6. then the width of flood plain which is contributing to the conveyance should be reduced. This would represent an unrealistic relationship between α1 and afflux h1. In a real river the flow on a wide rough flood plain with low flow will not have a uniform low velocity.5). It was concluded that the calculation procedure is still valid but the results from these cases are subject to greater uncertainty than for situations of unaffected downstream water level. Interaction of adjacent bridges 6. 6. viaducts. It is possible to reduce the calculated value of α1 by reducing the value of roughness on the flood plain used in the calculation of conveyance. 6.39 There are limits on the range of conditions of laboratory and field tests under which the USBPR method was developed.4 (the maximum value on Figure H2). 6.17. There are likely to be areas where there is zero or negative flow on the flood plain. the degree of variation in α1 is not too great and a division similar to that shown in the worked examples will give a reasonable value. The new bridge will be modern whereas the existing bridges may be of any type. the afflux for that bridge should be estimated from HR (1988) (Ref No. The results from applications of the method to situations outside this range should be treated with some caution. In this case it will be necessary to amend the cross section. This would lead to a large value of α2. recalculate the conveyance in accordance with paragraph 5.38 The value of α1 can be very sensitive to the values of roughness chosen to represent the main channel and the flood plains.2. Accurate estimates of the afflux caused by adjacent bridges of different types can only be obtained by physical modelling. such as a weir. Where one of the bridges is an arch bridge. and so the flow widths used in the calculations have effectively been over-estimated. for example standing water or large eddy rotations. which is somewhat arbitrary.37 In the calculation of α1 the same division of the section into subsections is used as for the calculation of conveyance.10). Subsections of the flood plain where the flow is shallow (eg less than 0.36 A worked example of the bridge afflux calculation is shown in Appendix I. This is because the expansion which occurs downstream of the upstream bridge and the contraction which occurs upstream of the downstream bridge interact. If they were moved further and further apart the afflux would tend towards twice the single bridge value minus the backwater effect between the bridges.

For cases (c) and (d) it is recommended that the method is used to obtain first order estimates of the afflux but further studies would be required. and not all of the flood plain will have flowing water. In this case the method can be used for a first estimate. Case (b) is a wide flood plan where the water level will vary significantly along the embankment. In both cases only a proportion of the left bank flood plain flow should be used in the calculations because some flow will occur between the road and the edge of the floodplain. Volume 4 Section 2 Part 1 HA 71/95 6/6 August 1998 .Chapter 6 Assessment of bridge afflux crossings. In case (c) two separate calculations are required for points A and B. but further investigations are required. In case (d) the approach conditions are very different from those used in the experimental work.

Volume 4 Section 2 Part 1 HA 71/95 Chapter 6 Assessment of bridge afflux Figure 6/1 Afflux calculation procedure August 1998 6/7 .

Chapter 6 Assessment of bridge afflux Volume 4 Section 2 Part 1 HA 71/95 Figure 6/2 Typical road crossing layouts 6/8 August 1998 .

2 If hydrographs at a number of locations along a reach of river are examined during a flood event. 7. and short but intense storms which frequently occur in the summer.3 The degree of attenuation which occurs depends on peakiness of the discharge hydrograph.1 During a flood event the discharge at a given point on the river will build to a peak and then reduce gradually. This effect is described as attenuation of the discharge hydrograph and is caused by water flowing out into storage areas on the flood plain rather than continuing downstream. for peaky hydrographs the filling of storage volumes can cause significant attenuation. the build up and recession is generally slow with a long period of flow at or near the peak flow to give a very flat hydrograph shape. It should be noted that storage has no effect under steady flow conditions. On small rivers the flood is likely to be a more rapid and shorter event thus producing a more sharply peaked hydrograph shape. 7. Peaky hydrographs occur on smaller catchments. This description is an over-simplification since there is likely to be additional inflow at locations within the reach. 7. the overall effect of flood plain storage is to reduce the peak discharge downstream. 7. A plot of this relationship is termed a discharge hydrograph. The method is very sensitive to the slope of the stage discharge curve at the peak flood discharge.4 Flat hydrographs occur on large catchments such as the Lower Severn. for example. the effects of any subsequent development on the flood plain will be additive and the cumulative effect of developments could therefore become significant. 7.Volume 4 Section 2 Part 1 HA 71/95 Chapter 7 Flood Plain Storage 7.5 It follows that if the available flood plain storage volume is reduced by construction of a road embankment then the attenuation will be less and the peak discharge downstream will not be reduced to the same degree. FLOOD PLAIN STORAGE Outline description of the problem 7. The shape of such a hydrograph can vary greatly. However. A method for calculating the increase in water levels due to loss of storage is given in paragraph 7.8 below. It is based on a method developed at HR in response to a specific problem. Recommended method for calculating water level increase due to loss of flood plain storage 7.6 As a rule. the effect of any one development on peak water level will be small. However. For flat hydrographs the potential storage volume will already have been nearly filled with water before the peak flow arrives and so the attenuation effect is very small. On large rivers. A higher peak discharge will consequently cause higher peak flood levels downstream. For this reason River Authorities often require compensatory flood plain storage to be provided.8 The increase in peak water level ( h) during a flood due to loss of flood plain storage can be calculated as follows: where = increase in peak water level ∆h (metres) ∆ Pa Β = loss of inundated area (m2) = mean inundated width for reach (m) (assumes small changes) August 1998 7/1 . which means that the peak of the flood is moving downstream as a wave. or on catchments where reservoirs or storage ponds are used to attenuate the hydrograph peak. it will generally be found that the time at which the peak flow occurs is progressively later at the next location downstream. Factors which increase the peakiness of a hydrograph include steepness of catchment. 7. an order of magnitude estimate is considered adequate for this application as the numbers are generally rather small. However. large urban area leading to rapid runoff.7 The method only provides an order of magnitude estimate (so that estimates may be in error by a factor of ten). However. where small uncertainties lead to large errors. It will also be found that the value of the peak flow tends to decrease as the flood wave moves downstream.

In order to calculate a stage/discharge curve it is necessary to obtain a cross sectional profile of the river and flood plain at the site where the road impinges on the flood plain. The water surface slope can be calculated as described in paragraph 5. These are often presented in the form of stage hydrographs and will therefore require conversion to a discharge hydrograph using a suitable rating curve. but to use the method for this purpose. See Note 3. 7. The peak design discharge of a given return period can be calculated as described in paragraph 4.If it is not possible to obtain a hydrograph shape as described above. This should be obtained from the nearest gauging station if possible. The hydrograph with the highest peak discharge should be used for this calculation. The stage/discharge curve is produced in the manner described in paragraphs 5. the rate of change of slope can be calculated.9 From estimates of the slope of the flood event hydrograph just prior to the peak and just after the peak. See Note 3 = water surface slope.dQp h = curvature of discharge hydrograph at the flood peak.10 ∆t might typically be 3600 seconds for a peaky hydrograph and 36000 seconds for a flat hydrograph on a large river. it will be necessary to calculate a suitable shape. Qp s = peak discharge (cumecs). but they should normally be adequate for this application. See Note 2. Note 4. This is equal to the slope of the stage/discharge curve for the river cross section at the flood peak.Chapter 7 Flood Plain Storage Studies Report. The recommended method for producing a flood hydrograph shape suitable for the calculation of curvature is the FSR Unit hydrograph method.∆t) + = Q at (peak + ∆t) Qp ∆t is in seconds 7. the calculation of hydrograph curvature requires a flood hydrograph shape. It should be noted that hydrograph shapes vary considerably.5. Note 3. Note 2. See Note 1 Volume 4 Section 2 Part 1 HA 71/95 7. from the River Authority or the UK Surface Water Archive. This is briefly discussed in paragraph 4.11 The impact of the loss of flood plain storage on peak river flow is: ∆Qp = ∆h.14. However.12.23. The value of ∆t should be selected to provide a reasonable definition of the curvature of the hydrograph peak. It follows that: = rate of change of water level with discharge.15 and 5. it will be necessary to refer to the Flood 7/2 August 1998 . These will only be available from gauging sites. Note 1. where Qp= Q at (peak .

straight river. 8. complexity and reliability. etc) simplified storage calculation when applying computational models because the results will only be as good as the formulae used to represent the road crossing. A computational model of the site can take account of flood plain flow and storage. 8.6 A physical model is likely to provide the most accurate estimates of the effects of the details of a site on flooding. A physical model demonstrates flood plain flow patterns and can be used to identify where flow and ponding occurs on flood plains. Care should be taken August 1998 8/1 .5 Taking into account the present state of computational modelling. - 8.1 The methods described in Chapters 4 to 7 provide first order estimates of the various effects that highway construction on flood plains may have on flood water levels. Simplifications include: hydrological analysis using limited data hydraulic analysis in which a single section is assumed to be typical of the river crude estimates of channel roughness and water surface slope bridge afflux analysis based on using ideal geometries (eg straight roads. More detailed studies may be required in the following cases: large potential for flood damage in the neighbourhood of the bridge or within the backwater length of the site flood plain viaduct required whose cost may be substantial compared with the costs of more refined analysis or modelling complex local river geometry or several structures in close proximity approach embankments severely skew to the axis of the river valley. regular flood plain.2 above. FURTHER STUDIES 8. It can be used in the design of the road crossing to identify the optimum locations for flood plain openings including bridges and culverts. A physical model is essential for the hydraulic design of road crossings on wide flood plains. 8. but their limitations must be appreciated. and the models have no systematic way of identifying which parts of the flood plains have flowing water and which parts are ponded during a flood.Volume 4 Section 2 Part 1 HA 71/95 Chapter 8 Further Studies 8. account must be taken of possible scale effects in interpreting the measurements. and is normally carried out computationally.2 If the methods indicate that the effects will be significant. a physical model would generally be recommended for investigating bridge afflux problems of the types described in paragraph 8.3 • • • These options are arranged in probable ascending order of cost. a more rigorous approach will be required. However.4 A backwater calculation provides a steady flow water surface profile along a river. The methods are simplified to facilitate manual computation. Flood plain flow is poorly represented. skew angle not exceeding 45°. Further assessments are most likely to be: backwater calculation computational model physical model - - 8. The development of two and three dimensional models of rivers and structures may produce satisfactory computational techniques although it will be difficult to calibrate such models and select suitable discharge coefficients for new structures.

*** “Afflux at arch bridges”. * 12. 9. (1934). US Bureau of Public Roads (1978). Kindsvater C E. McGraw Hill. 1985. John Wiley and Sons. 5 volumes. Material from the “Hydraulics of Bridge Waterways” is used with the permission of the US Department of Transportation Federal Highway Administration. *** “Hydrometric Register and Statistics 1981-85”. Instn. Chichester (1978). USGS open-file report. 14. New York (1959). Institute of Hydrology. (1958). Techniques of Water-Resources Investigations of the USGS. 11. 1985 (August) and SR 182. US Dept. Report SR 60. *** “Backwater lengths in Rivers”. Civ. Acknowledgements 8. 6. ** “Bridge piers as channel obstructions”. * “Open Channel Hydraulics”. French R H. Design Series No. Pt. ** “Computation of backwater at open-channel constrictions”.16: The FSR rainfall-runoff model parameter estimation equations updated”. ** “Weirs and Flumes for Flow Measurement”. 16. British Geological Survey. 13. Natural Environment Research Council.87. etc Available from technical libraries Difficult. * 2. Institution of Water and Environmental Management (IWEM). Availability *** ** * Readily available . Part 7. Technical Bulletin No. (1953). McGraw-Hill (1985). Structures (1991).booksellers. * “Open Channel Hydraulics”. August 1998 9/1 . “Measurement of peak discharge at width contractions by indirect methods”. Carter R W & Tracey H J.442. Perkins J A and Harrison A J M . Vol. HR Wallingford (1985 & 1988). of Transport. Natural Environment Research Council (1988). eg British Library 7. Hydraulic Design Series No. * “Flood Studies Supplementary Report No. Chow V T. Cragwall J S. 4.FHWA-1P-85-15.Volume 4 Section 2 Part 1 HA 71/95 Chapter 9 References 9. Engrs.5. Matthai H F.2. USGS Circular 284. USGS Geological Survey (1977).2. Report No. *** 3. ** “Flood Studies Report”. * “Hydraulics of Bridge Waterways”. pp571-582 (1989). 1988 (December). White W R. but may be obtained at specialist libraries. Samuels P G. “Water Practice Manual”. 15. Yarnell D L. (1967). 5. ** “Hydraulic Design of Highway Culverts”. REFERENCES 1. River Engineering (1989).* “Computation of backwater and discharge at width constrictions of heavily vegetated flood plains”. Ackers P. Part 8. (1985). US Department of Transportation. Federal Highway Administration. “Computation of peak discharge at contractions”. US Department of Agriculture. Proc. 10. Material from the “Flood Studies Report” is used with the permission of the UK Institute of Hydrology. Natural Environment Research Council (1975).

ENQUIRIES All technical enquiries or comments on this Advice Note should be sent in writing as appropriate to:- Head of Division Road Engineering and Environmental Division St Christopher House Southwark Street London SE1 OTE N S ORGAN Head of Division The Deputy Chief Engineer The Scottish Office Industry Department Roads Directorate New St Andrew's House Edinburgh EH1 3TG N B MACKENZIE Deputy Chief Engineer Head of Roads Engineering (Construction) Division Welsh Office Y Swyddfa Gymreig Government Buildings Ty Glas Road Llanishen Cardiff CF4 5PL B H HAWKER Head of Roads Engineering (Construction) Division Assistant Chief Engineer (Works) Department of the Environment for Northern Ireland Roads Service Headquarters Clarence Court 10-18 Adelaide Street Belfast BT2 8GB D O'HAGAN Assistant Chief Engineer (Works) August 1998 10/1 .Volume 4 Section 2 Part 1 HA 71/95 Chapter 10 Enquiries 10.

The capacity of the main river channel when the water level is at the top of the banks. etc). flood plain. A hydrological model of a river catchment used to predict flood flows. A method of calculating the steady flow backwater surface profile along a river The length of river affected by the backwater of a structure. rubbish etc. in which the catchment is divided into sub-catchments. river and flood plain. The largest flood which occurs in a year. branches. Storage which is provided for flood water as part of a scheme to compensate for the volume which is lost as a result of the scheme (for example. The process of matching predicted water levels or flows to observed water levels or flows in the hydraulic analysis of a particular flood event. The carrying capacity of a channel. Increase in upstream water level caused by the presence of a bridge.Volume 4 Section 2 Part 1 HA 71/95 Appendix A GLOSSARY AND ABBREVIATIONS Glossary Abutment (bridge) Afflux Annual flood Annual maxima (of river flows) Attenuation (of a hydrograph) Backwater Support for the end of a bridge or viaduct. A general term used to describe a river. caused by water passing into storage. Flow predictions are made for each sub-catchment and combined using a flow routing technique. by the construction of embankments. Material collected by a river in flood. or other open channel. The water level increase caused by a structure above the water levels which would occur if the structure was not present. The reduction in height of a flood hydrograph as it moves along a river. The area from which rainfall runoff flows into a river. Backwater calculation Backwater length Bankfull flow Calibration Catchment Channel Debris Compensatory flood plain storage Conveyance Distributed catchment model Flood hydrograph August 1998 A/1 . A time series of Annual Floods. A graph of river flow against time for a river flood event. including trees.

The net 1-day rainfall of 5 years Return Period less the Soil Moisture Deficit. A measure of flow depth in a channel defined as the flow area divided by the wetted perimeter. A curve relating flood discharge to probability of occurrence and Return Period. Residual Soil Moisture Deficit (RSMD) Return period (for a particular river flow) A/2 August 1998 . The authority responsible for trunk road construction. Relationship between stage and discharge at a particular location in a river (or stage/discharge curve). A statistical analysis of flow data which is used to produce a relationship (of river flow data) between river flow and return period. The flood which has a 50% chance of being exceeded in any given year. A graph of flow (or water level) against time. The time period during which a particular river flow is equalled or exceeded once on average. The analysis of the passage of flood hydrographs down a river system. Culverts through an embankment which crosses the flood plain.Appendix A Glossary and Abbreviations Volume 4 Section 2 Part 1 HA 71/95 Flood plain Flood plain culverts Flow routing Frequency analysis The land adjacent to the main river channel which is liable to flooding. A place where river flow is measured on a regular basis. The main river channel in a river with flood plains. as follows: England Wales Scotland Northern Ireland The Department of Transport. Highways Agency The Welsh Office The Scottish Office The Department of the Environment for Northern Ireland Gauging station Growth curve Hydraulic radius Hydrograph Main channel Mean Annual Flood Overseeing Organisation Pier Rating curve Support for the spans of a bridge or viaduct.

streams and drainage ditches conveying drainage water. Skew Soil moisture deficit Stage Statistical method (for river flows) Steady flow Sub-catchment Flow which does not vary with time. A measure of the proportion of rainfall which is absorbed into the soil and therefore does not runoff directly into the surface drainage system. Percentage Runoff is runoff expressed as a percentage of the rainfall. The network of rivers. as follows: England and Wales Scotland Northern Ireland The National Rivers Authority Internal Drainage Boards River Purification Authorities Islands Councils The Department of Agriculture. The proportion of rainfall which runs directly into the surface drainage system. River water level. The slope of the water surface along the length of a river. Surface Drainage System Unsteady flow Water surface slope Abbreviations FSR MAF CH FP DMRB u/s d/s Flood Studies Report (NERC) Mean Annual Flood (m3/s) Channel Flood plain Design Manual for Roads and Bridges (HA) upstream downstream August 1998 A/3 . A method of analysis based on a statistical analysis of river flow data. A division of a Catchment which is drained by a tributary or particular part of the river system. Flow which varies with time.Volume 4 Section 2 Part 1 HA 71/95 Appendix A River Authority The authority responsible for rivers. Northern Ireland Roughness Runoff The resistance (of a channel) to flow. Where a road crossing is not at right angles to the river.

Volume 4 Section 2 Part 1 HA 71/95 Appendix B NOTATION A ai Ac B B b bs C D F f g Hu Hd h h1 i J K L M n ni P Pa Pi Q Qp qi R Ri s T V vi Wp y κ κb ∆κp ∆κe ∆κb φ Total cross sectional area of river Area of segment i Area of culvert Average flow width in culvert Mean inundated width of flood plain Width of construction Width of construction of a skew crossing measured along the centreline of the road Regional coefficient Bankfull depth Non-exceedance probability Drowning coefficient Acceleration due to gravity Upstream water level above culvert invert Downstream water level above culvert invert Water level Afflux Rank number or segment number or subsection number Pier blockage ratio. as defined on Figure 5/4 Conveyance Backwater length Bridge opening ratio Mannings roughness coefficient Value of ‘n’ for segment i Wetted perimeter Inundated area of flood plain Wetted perimeter of segment i Discharge Flood peak discharge Discharge for subsection i Hydraulic radius Hydraulic radius of segment i Water surface slope Return period Average velocity at river cross section Average velocity for subsection i Width of pier normal to flow Reduced variate Kinetic energy correction coefficient Overall backwater coefficient Base curve coefficient Pier blockage incremental coefficient Eccentricity incremental coefficient Skew incremental coefficient Skew angle (degrees) α August 1998 B/1 .

Note 2.see Note 6 Proportion of the catchment area which is covered by lakes Data for these parameters are to be found in Appendix D: Note 1. This can be a time consuming and laborious process for large catchments. The boundary of a catchment is drawn in by judging the line of the watershed between drainage systems ie the highest point between basins which flow in different directions.see Note 3 Fraction of the catchment area which is urban. It should be noted that the catchment area of gauging stations is given in the Hydrometric Register and Statistics (NERC. MAF = 0. If an approximate value is used it is important to ensure that the resulting MAF value is checked against another estimate from gauging station data. ie the correction factor discussed in this section is of increased significance. August 1998 C/1 .03S10850.see Note 2 Number of junctions of streams per km2 catchment area .94STMFRQ0.Volume 4 Section 2 Part 1 HA 71/95 Appendix C Statistical Method for Calculating MAF APPENDIX C STATISTICAL METHOD FOR CALCULATING MAF The following equation should be used for the Essex Lee and Thames area shown in Appendix D.23 RSMD1.see Note 4 Residual Soil Moisture Deficit (mm) . Note 3. In order to obtain a first estimate of MAF.see Note 5 Slope of the main stream (m/km) .27SOIL1.85] where C SOIL RSMD S1085 LAKE = = = = = Regional coefficient . 1988). The catchment area AREA can be calculated sufficiently accurately from commercially available maps at scales of about 1:50000.see Note 1 A measure of soil types in the catchment with characteristic runoff potential . Figure D1. In all other cases.5 where AREA STMFRQ URBAN = = = Catchment area (km2) . The Regional Coefficient C is read off the map shown in Figure D1.16(1+LAKE)-0.373 AREA0. The stream frequency STMFRQ is to be measured on commercially available maps at scales of about 1:25000. the following equation should be used: MAF = C[AREA0. it may be helpful to use the mid value of STMRQ from the appropriate range given for different areas of the UK in Figure D3.70STMFRQ0.52 (1+URBAN)2.

.5 S5) S1 + S2 + S3 + S4 + S5 where S1. The mid value of the range of slopes appropriate to the area should be used. Note 6. failing better information about the river under consideration.15 S1 +0. C/2 August 1998 .. Note 5. The soils are classified according to runoff potential (RP). The residual soil moisture deficit RSMD is read off Figures D6a-c. an approximate value of S1085 can be selected from the map shown on Figure D4. The slope between these two points is then calculated to give S1085. The slope of the mainstream S1085 is assumed to be calculated in accordance with the method described in the Flood Studies Report. In the first instance.Appendix C Statistical Method for Calculating MAF Note 4.S5 denote the proportions of the catchment covered by each of the soil classes 1-5 as shown on Figures D5a-p. This requires the measurement of the largest stream length from the catchment outfall (in this case the gauging station) to the furthest point upstream.40 S3 +0. A single average value for the catchment area is required to be calculated. The parameter SOIL is calculated as follows: Volume 4 Section 2 Part 1 HA 71/95 SOIL = (0. as marked on commercially available maps at scales of about 1:25000.30 S2 +0.45 S4 +0. . Points at 10% and 85% along this main stream length are identified and the elevation of these points noted.