Frequently asked questions

Registering with us is completely free. To register for a free account, go to the [LINK] and complete the various fields. Once you have completed the form, the site will give you access to the secure area and an email with your username will be sent to your registered email address.

If you register for a free account, you will be able to:

• Access all tools for free
• Generate pdf reports for specific tools
• Receive latest news about SuDS and training

To register for an account on the UK SuDS website, you only need to supply us with your full name, company or organisation name and type, email address. You will also need to enter the username and password of your choice.

Registered users who wish to amend or close their account need to contact the UK SuDS team by email at uksuds@hrwallingford.com. Once their account is deleted, all the data they may have stored in their members' area will also be deleted. 

The attenuation storage volume is based on the simple assumption of 100% runoff from paved surfaces and no runoff from pervious areas (in line with Sewers for Adoption 7th ed.) and this is reasonably valid for sites with a high proportion of hard surfaces. This can be shown to be reasonably equivalent to the results produced by the New PR equation used currently by the water industry in drainage modelling. 

Currently there is a practice of designing drainage systems using 84% runoff coefficient from paved surfaces only. This has been derived from the original Wallingford Procedure (1983) runoff model (the Old PR equation). It is not a conservative assumption when applied to high intensity rainfall and this practice should only be used if it is reasonable to assume that no runoff from pervious areas take place. (The Old PR equation was replaced by the New PR equation in 1991 because it under-predicted runoff from large or long duration events).

Drainage engineers who carry out simulation modelling of extensive sewerage drainage systems  tend to use more refined runoff models (the New PR equation and the latest equation referred to as the UKWIR equation). These assume runoff coefficients for all surface types which vary with rainfall depth and other parameters. These models can be used by those familiar with their use in tools which incorporate them, but they are rarely used for site development drainage.

Current practice uses formulae for calculating greenfield runoff which do not have any gradient terms in them. This results in lower greenfield flow rates than probably take place in practice in most circumstances. For the developer this implies unnecessary cost, while for the Planners and Environmental Regulator it provides a margin of safety for controlling downstream flooding. It is important to recognise that the type of event which actually gives the peak flow rate on a site is of relatively little importance in the receiving catchment and it is the longer less intense storms which are more relevant.

What is not acceptable when using the UKSuDS.com tool is the arbitrary modification parameters such as soil type to produce higher rates of flow for the site to compensate for issues such as site gradient.

In the long term there needs to be research which focuses on the change in volume of runoff together with an assessment of the impact on the receiving water course to arrive at a new set of drainage criteria for site runoff. 

Climate change is believed to result in higher intensity extreme rainfall events. The approach for making an allowance for future conditions is to use the recommended greenfield runoff equations, or previously developed analysis for present day conditions, and then factor the design rainfall by an appropriate value (usually now 40%) to assess the storage and conveyance requirements for the site.

It should be noted that using present day rainfall and factoring the resulting storage volume by 40% will significantly under-predict the attenuation storage required.

Greenfield runoff volume is calculated for the 1:100 year 6 hour event as this volume is used in the design of the drainage system . The volume is a very approximate greenfield runoff estimate which assumes that the proportion of runoff is equal to the SPR value for the soil. This means a coefficient of runoff between 0.1 and 0.6 is assumed. Although values for some soil classes have SPRHOST which are lower than 0.1, the use of these values is not advised.

The document [LINK] (SC030219 rev. E 2013) should be referred to for more information on the basis of using SPR values for estimating greenfield runoff volumes.

The ReFH2 plot scale tool also has the capability of predicting the greenfield runoff volume for any specific event. As it calculates a different wetness factor to that of SPRHOST, the greenfield volume calculated will be different.

The 1:100 year 6 hour rainfall depth is relatively easily established. However due to the basis of this criterion being a very approximate justification, a depth of 60mm can be considered as a suitable approximation to use if the actual depth is not known. Whether a climate change uplift factor is used is a matter for agreement, but is not advised. The arbitrary nature of the criterion does not really justify such an increase in this design criterion being taken on this aspect. Any modification of the design rainfall depth should probably be more associated to the sensitivity of the receiving catchment and the distance downstream that the site might have influence in terms of increasing the flood risk.

The Greenfield runoff rate which is to be used for assessing the requirements for limiting discharge flow rates and attenuation storage for a site should be calculated for the whole development area (paved and pervious surfaces - houses, gardens, roads, and other open space) that is within the area served by the drainage network whatever size of the site and type of drainage system. Significant green areas such as recreation parks, general public open space etc., which are not served by the drainage system and do not play a part in the runoff management for the site, and which can be assumed to have a runoff response which is similar to that prior to the development taking place, can be excluded from the greenfield analysis.

The exclusion of large green areas is important when using the UKSuDS.com storage assessment tool as it is only valid for use where the paved proportion of the site is larger than the pervious proportion (PIMP is greater than 50%). Ignoring for a moment the issue of the over-extrapolation and validity of the approach, calculating a greenfield discharge rate for the whole site including this type of public open space, will result in a higher Qbar value than if it is excluded. If this is subsequently applied to the (smaller) area served by the drainage system, then the storage volume estimation will be under-sized.

Where green spaces cannot be excluded because they have been landscaped to alter their runoff characteristics or to assist in managing the runoff from extreme events and this results in a PIMP value which is less than 50%, then the use of the UKSuDS tool is not valid and other more appropriate runoff modelling and management techniques should be used for assessing the storage requirements. If it is applied it should be recognised that the estimated storage volume is under-predicted. The reason why the tool should not be used is that no runoff is assumed to take place from pervious areas in the development and this is partially compensated by using 100% runoff from paved areas (in line with sewers for Adoption). 

The tool has been known to be used by applying only the paved area as being the site area. Applied in this way it is implying that all the vegetated areas of the site (gardens etc.) continue to discharge as they did prior to development. It will also result in small values of Qbar. This is a very conservative position, but is an approach that could be taken if it was thought to be appropriate for a particular situation. 

The SPRHOST value can be obtained from other sources such as

• the FEH web service, 
• Soil maps in conjunction with IH126, 
• the ReFH2 CD ROM tool or 
• the Cranfield University National Soil Resources Institute (LandIS).

The IH124 and FEH statistical equations are derived from measured river flows based on catchments all of which are larger than 25km2. The ReFH2 method is also a catchment scale model, but with a modification for use with plot scale analysis. The plot scale work was calibrated against a limited data set of sites with relatively moderate sized rainfall events, and there is relatively little information provided on this work in the “The Revitalised Flood Hydrograph Model ReFH2: Technical guidance (2015)”.

The application of these methods to sites which are often less than 1ha is an extrapolation which is well beyond what would normally be accepted. The official position on the use of the equations is to limit them to 50ha, (which is why ReFH2 plot scale is applied initially at 50ha), but their accuracy at even this scale is questionable. The official advice for estimating site development flow rates is to calculate the greenfield flow for 50ha and extrapolate the value linearly by proportion to the size of the site. As the AREA term in both equations are to a power slightly less than 1.0, this means that flow rates are slightly smaller when calculated in this manner, and therefore they are conservative for the purpose of setting discharge limits. This approach has been applied in www.UKSuDS.com.

There have been suggestions that as these formulae are catchment based, that the flow rates for the site should be based on a linear interpolation of the flow for the catchment in which it is located. This is convenient for those who use the ReFH2 tool as catchments are pre-defined in the tool and peak flow rates and growth curves can be derived easily. However the counter argument is that soil classes can vary considerable across any catchment, and the average values used for this and other parameters do not reflect the specific runoff characteristics of the development. 

However the strongest argument has been that many small organisations do not have access to the ReFH2 tool and that as all methods are of limited accuracy at the site scale, any of the methods can be used to avoid having to purchase the software package. Until there is free or low cost access to such tools it is still considered appropriate to use the other recommended methods.

The general position taken on minimum flow rate in the past has been to use 5l/s to avoid throttle sizes from being too small and thus create a risk of blockage. However as most development sites are small, this flow rate often does not really provide the degree of throttling of flow that is needed. Although there is little evidence available on blockage risk of drainage systems, as the DCG allows adoption down to a 50mm orifice, it is now generally felt that 2 l/s (which can be achieved with a vortex control regulator), is now an appropriate target for the minimum flow rate. This does not mean that lower flow rates might be reasonably asked for in special circumstances. However as blockage risk is very much a function of both the orifice size and the solids (twigs etc) transported in the surface water runoff, it is also now advised that a risk assessment of blockage is made for all control orifices less than 100mm in diameter. This should lead to appropriate selection of SuDS design which minimises the amount of solids (debris and sediment) in the flow. 

HOST, i.e. Hydrology Of Soil Types, classifies the soils of the United Kingdom into 29 categories. These classes are based on a series of conceptual models that simulate the hydrological behaviours associated with the soils and it interprets soils physical properties and their effects on the development of soil water.

SOIL indices (1 to 5) are defined in the Flood Studies Report (NERC, 1975). The index broadly describes the maximum runoff potential and was derived by a consideration of soil permeability and topographic slope.

SOIL type 1 is sandy highly permeable material with permeability reducing as the SOIL value increases. SOIL type 4 is heavy clay and 5 (which is rarely applied) is exposed rock. For each of these categories there is a value of SPR which is used in the IH124 equation. These were 0.1, 0.3, 0.4, 0.47, 0.53 respectively. The minimum value for SPRHOST used in this tool is constrained to the minimum for SOIL at 0.1  as this is the minimum value used when the IH124 equation was developed.

The return periods that are traditionally set for drainage design criteria are:

• 1:30 years: roads should not flood for events up to this magnitude
• 1:100 years: runoff discharged from the site should be managed on site and discharged to meet a stipulated flow rate.

In addition to these, there are a number of additional criteria which are normally used, including:

• 1:1 and 1:10 years: Lower return periods than the traditional 1:30 years are considered where flooding can be tolerated in specific locations;
• Where SuDS are being designed for water quality performance, there are a range of methods that are used, but it is quite common to assess their hydraulic characteristics using a 1:1 year event.
• 1:100 years: safe access / egress on major roads in locations such as floodplains;
• 1:100 years: no flooding of property or critical infrastructure, with a suitable freeboard specified which should relate to the uncertainty associated with achieving that freeboard;
• 1:100 years: As stated above, the management of site runoff should be for it to remain on site for a period of time until it can be discharged off site. However flood flows can pass across the site dictated by the topography or pond at low points or due to impediments to the flow (properties). In addition, off site flooding may come onto the site and this should be managed as well.
• 1:200 or greater: Locations adjacent to the sea (and rivers in Scotland) use higher return periods to take into account the greater consequence of flooding from such sources.
• Extreme return periods: There are always exceptions where higher return periods are to be considered. Nuclear sites or locations in the flood route of dams located upstream and so on, require consideration of more extreme events. Some very high value property may also warrant higher levels of service.

it is important to recognise that design criterion can be relaxed as long as the implications are clearly understood. For instance car parks could be allowed to flood to specific depths for short periods of time for a return period which is considered to be acceptable as long as at the 1:100 year event damage is minimal and there is no risk to health and safety. Recreational areas can similarly be used to assist in managing flooding to prevent excessive construction costs of underground storage systems.

In designing or assessing the performance of drainage systems, the critical duration event must be considered and that this duration storm will vary depending on the drainage element being considered.

The simple assumption is normally made that the return period of the rainfall is the same as the design level of protection against flooding. This is reasonably valid where the area is dominated by paved surfacing and there is limited influence from either  uphill areas or downstream water levels.

The Winter rain acceptance potential map for the UK, produced by the Wallingford Procedure is provided within the Wallingford Procedure data available from HR Wallingford's [LINK].

Alternative maps are available from the original FSR document. There is a soil category referred to as urban on the FSR map that is used for city areas. It is assumed that the SOIL value is 4 for the purposes of storage analysis.

There are many formulae that can be used for calculating greenfield runoff, however there are only three methods which are regarded as being acceptable for being used for planning applications for drainage design or other similar purposes. These are:

• The equation derived in IH 124, Flood estimation for small catchments. (1994); 
• The FEH statistical equation by Kjeldsen et al. (2008); and the recently produced
• ReFH2 plot scale equation by CEH (2015).

The IH124 equation predicts a greenfield peak flow rate estimate for the mean annual flood called Qbar (a return period of approximately 1:2.3 years). This value is then factored by a growth rate parameter to give a flow rate for other return periods.

IH124

Qbar = 0.00108 x (AREA)0.89 x SAAR1.17 x SPR2.17
where:

• Qbar is the mean annual flood flow from a rural catchment (m3/s)
• AREA is the area of the site (km2)
• SAAR is the standard Average Annual Rainfall for the period 1941 to 1970 (mm).
• SPR is Standard Percentage Runoff coefficient for the SOIL category.

Information for SAAR and SOIL are available from paper maps which were issued by the Flood Studies Report (FSR 1975) and also subsequently in 1983 in the Wallingford Procedure.

The value for SAAR was reassessed in the Flood estimation Handbook (FEH 1999) as parameter AAR for the period 1961 – 90, and then again in the reissue of FEH in 2015 (called FEH13) with the period extended through to 2010. The differences between any of these values is small for most location across the country. The advantage of using the digitally based information is the increased level of accuracy for a specific site. However this must be seen in the context of the derivation of the formula which was produced prior to the digital age, which means that the increased resolution of any parameter does not necessarily mean the use of the equation is necessarily any more accurate. 

Similarly the value of SPR for the 5 soil categories in the Winter Rainfall Acceptance Potential (WRAP) map can be substituted for digitally based information of parameter SPRHOST based on a greater differentiation of 29 soil types based on IH126 Hydrology of soil types: A hydrologically based classification of soil types of the United Kingdom, (1996) if it is available. Unlike the parameter AAR, the two values of SPR and SPRHOST are not exactly the same measure of the hydrological characteristics of soil. However the difference is considered to be sufficiently small as to justify the substitution of the parameter value based on the advantage of being much higher resolution soil information. 

Appendix 5 of the Environment Agency document “Preliminary rainfall runoff management for developments”, SC030219 ed. E gives an analysis comparing the two equations based on the substitution of the SPR parameter. As a result of the analysis reported in Appendix 5 of SC030219 ed. E although SPRHOST values are suggested as being used to substitute for the 5 soil types in the IH124 method, it is recommended that a lower bound limit to the value of 0.1 is used.

Unfortunately there are two sets of SPRHOST values available; those used by FEH and those related to the “predominant soils” classification. The values are very similar for most soil types, but a few have more significant differences. The analysis summarised in Appendix 5 of SC030219 indicates that the latter values would seem to be more appropriate, but a fundamental comparison against river flow data has not been made.

FEH statistical equation

The FEH statistical equation predicts the median annual flood called Qmed (a return period of 1:2 years).

Qmed = 8.3062 x (AREA)0.851 x 0.1536(1000 / SAAR) x FARL3.4451 x 0.0460(BFIHOST x BFIHOST)

where: 

• Qmed is the median annual flow rate; the 1:2 year event. (m3/s)
• AREA is the area of the catchment (ha).
• AAR is the standard average annual rainfall for the period 1961 to 1990 (mm).
• FARL is a reservoir attenuation function and is set at 1.0 and therefore has effectively been ignored. This means that areas which have water bodies which attenuate the runoff will over-predict the greenfield runoff rate.
• BFIHOST is the base flow index derived using the HOST classification.

As stated above, In contrast to the FSR based methods on which IH124 is based, the FEH and ReFH methods are all based on digital and higher resolution information. The formulae are generally regarded as being more accurate than the IH124 though this can only be assessed at the catchment rather than site scale.

It is interesting to note that analysis for the peak flow formula found the use of BFIHOST (a base flow index) to produce a better correlation than the use of SPRHOST. As flood flows are not intuitively linked to river base flows this provides a warning to treat all these formulae with a degree of caution in the prediction of high runoff flow rate values. 

To obtain the flow rate for any specific return period, a growth factor based on curves given in FSSR 14 (Review of Regional Growth Curves, 1983) can be applied to both the FEH and the IH124 equations if the ReFH2 tool is not available to provide growth curve information. Use of the ReFH2 tool does require appropriate pooling selection of representative catchment data to provide a relevant growth curve.

ReFH2 plot scale equation

The latest release of the ReFH2 software has a module specifically developed for plot-scale application of the method. This method uses specific plot-scale equations for Time to Peak (Tp) and Baseflow Lag (BL) which are different to the catchment scale tool. The areal reduction factor for the rainfall is set to 1.0 and the winter design storm profile is recommended.

Tp is the unit hydrograph time to peak in hours and for plot scale applications is calculated by:

• Tp = a PROPWETb AREAc (1 + urbext2000)d SAARe
• BL is the baseflow recession constant or lag in hours
• BL = a PROPWETb AREAc (1 + urbext2000)d BFIHOSTe

Tp and BL are evaluated by the software for a catchment area of 0.5 km2. For greenfield sites, the urbext parameter will be set to zero. These values are then used to compute the design storm duration which is then applied to the site. The plot-scale models use AREA as an alternative to the catchment method descriptor DPLBAR (mean drainage path length), and also SAAR as an alternative to the catchment descriptor DPSBAR (mean drainage path slope). This is because DPLBAR and DPSBAR cannot be estimated at the plot-scale by the FEH catchment descriptor software. Because the formulae are applied at 50ha to define the design storm duration, the length of the event is often significantly longer than event durations calculated by other methods at site scale. This results in a lower intensity storm and therefore a lower predicted runoff peak flow rate than might give the actual peak runoff rate for a site. The recommendation for using a winter profile event, when typical events causing maximum peak runoff are summer events, is unclear.

As this method is very new it is recommended that comparisons are normally made with other methods to provide comparative information on predicted greenfield runoff rates.

Also, as the method uses a hydrograph approach, there is scope for using this output to evaluate site storage requirements.

Designing for exceedence is the assessment of the impact of a larger event than that used for designing the drainage components on the site. There is no prescription usually for what the assessment return periods should be, though embankments involved in storing significant volumes of water should be assessed using guidance associated with management of reservoirs.

The method of analysis requires information on finished ground levels, floor levels, structures which would be impediments to flows (buildings, walls) and possibly an assessment of the risks of future changes on the site which might create an obstacle to flows. The assessment can range from indicative flow routing using the level and structures information through to detailed 2D modelling. Depth of flooding is useful to predict related to return period and related to critical levels such as floors and any critical infrastructure. In addition it is useful to establish the duration of flooding especially at locations which might be planned for temporary flood storage if the design criterion is not particularly infrequent. For instance a car park might flood for return periods greater than a 1:5 year return period. The depth duration and location of flooding would be important to establish in this situation.

Flows generated on the site and also flows which might come onto the site from other areas need to be considered.

In addition to assessment of extreme rainfall conditions generating such flows, it is also important to assess flooding from other causes. These include failed drainage systems (blocked outlets, collapsed pipes) and even burst pressure mains, particularly where a water supply main within a site is reasonably large.

Greenfield runoff is usually calculated as the peak rate of runoff for a specific return period due to rainfall falling on a given area of vegetated land. There are methods which predict only the peak rate of flow while others produce a runoff hydrograph.

Although greenfield runoff assessment is often a requirement for development sites, the approved formulae have all been derived from river catchment runoff data. The recent publication of ReFH2 and the plot scale method has been produced to try and make the predictions more relevant and accurate for site developments. 

The extrapolation of these formulae to site scales is justified on the basis of providing consistent approaches that can be universally applied, rather than giving an accurate assessment of the actual runoff rate from a site. No method would claim to provide an accurate assessment of site scale runoff. 

There are a number of factors which are not taken into account which can result in under-prediction of runoff rates from a site. These include:

• The vegetated land characteristics (whether it is treed, grassland, farmed or otherwise landscaped) is not used;
• The gradient of the site is not used;
• Rainfall intensity is not used.

However it must be stressed that although the formulae claim to give a greenfield flow rate, and that it might be much less than it might be, this does not invalidate the runoff control requirements to protect the catchment downstream.

Every drainage unit on a site, if being designed using standard design rainfall events, will have a critical duration which will result in the maximum volume of storage or velocity. It is important to run a range of duration events to ensure the worst case condition is found for each drainage element on the site.

The Flood Estimation Handbook (FEH) 2013 rainfall data is the industry standard estimation of design rainfall. 

The data can be found on the FEH Web Service at https://fehweb.ceh.ac.uk/. The FEH Web Service is a product of the Centre for Ecology & Hydrology (CEH), and is developed, maintained and made available by its agent Wallingford HydroSolutions (WHS).

Long Term Storage is the term given to the volume of temporary storage which needs to be provided for the additional volume of surface water runoff that is generated by the development that is greater than the volume of greenfield runoff. The greenfield runoff volume is calculated using the 1:100 year 6 hour event. This volume is the amount that can be discharged at the 1:100 year greenfield runoff rate.

The additional runoff volume should be discharged from the site at a flow rate less than 2l/s/ha for this event. As critical duration events for the design of the site storage system will be much longer than 6 hours, the Long Term Storage volume is not calculated using the 1:100 year 6 hour event, but needs to be assessed using the critical duration event.

Qbar is the peak rate of flow from a catchment for the mean annual flood (a return period of approximately 1:2.3 years). It was produced in report IH 124, Flood estimation for small catchments. (1994) and is largely based on the work of the Flood Studies Report (1975).

Controlling flow rates from sites is currently based on Qbar for deriving the discharge rates for the 1:1 year and 1:100 year return periods. When the statistical FEH equation is used to derive Qmed, a conversion factor is applied to obtain Qbar.
More information can be found in “What formulae are used in calculating the greenfield runoff rate?”

Qbar = 0.00108 x (AREA)0.89 x SAAR1.17 x SPR2.17
where:

• Qbar is the mean annual flood flow from a rural catchment (m3/s)
• AREA is the area of the positively drained site (km2)
• SAAR is the standard Average Annual Rainfall for the period 1941 to 1970 (mm).
• SPR is Standard Percentage Runoff coefficient for the SOIL category.

Qmed is the peak rate of flow from a catchment for the median annual flood (a return period of 1:2 years). It is the key parameter used by the FEH statistical equation and ReFH2 in deriving river flow rates.
FEH statistical equation
Qmed = 8.3062 x (AREA)0.851 x 0.1536(1000 / SAAR) x FARL3.4451 x 0.0460(BFIHOST x BFIHOST)
where:

• Qmed is the median annual flow rate; the 1:2 year event. (m3/s)
• AREA is the area of the catchment (ha).
• AAR is the standard average annual rainfall for the period 1961 to 1990 (mm).
• FARL is a reservoir attenuation function and is set at 1.0 and therefore has effectively been ignored. This means that areas with water bodies which attenuate the runoff will over-predict the greenfield runoff rate.
• BFIHOST is the base flow index derived using the HOST classification

and uses parameters which are provided on paper-based information. In contrast the FEH based methods are all digital. 

The Flood Estimation Handbook, produced by the Institute of Hydrology in 1999 effectively replaced Flood Studies Report in the UK. 

The following QMED equation is from a revision to the statistical approach in 2008 (Kjeldsen et al, 2008).

Qmed = 8.3062 x (0.01 x AREA)^0.851 x 0.1536^{(1000 / SAAR)} x FARL^3.4451 x 0.0460^{(BFIHOST x BFIHOST)}, m³/s

where:

• Qmed is the median annual flow rate; the 1:2 year event.
• AREA is the area of the catchment in ha.
• SAAR is the standard average annual rainfall for the period 1941 to 1970 in mm.
• FARL is a reservoir attenuation function and is set at 1.0 and therefore has effectively been ignored. This means that areas with water bodies which attenuate the runoff will over-predict the greenfield runoff rate.
• BFIHOST is the base flow index derived using the HOST classification.

The Institute of Hydrology carried out a number of studies on revising the runoff equations produced in the original Flood Studies Report (1975). IH124 was specifically produced to address the runoff from small catchments. (Institute of Hydrology, 1994).

Although shown to be slightly less accurate than more recent FEH based methods, it is still considered to be an acceptable approach for assessing greenfield runoff rates. The IH124 equation estimates Qbar with the following equation:

Qbarrural = 0.00108 x (0.01 x AREA)0.89 x SAAR1.17 x SPR2.17, m3/s

Where:

• Qbarrural is the mean annual flood flow from a rural catchment (approximately 2.3 year return period).
• AREA is the area of the catchment in ha.
• SAAR is the standard average annual rainfall for the period 1941 to 1970 in mm (SAAR 41-70). SAAR 61-90, which was analysed for FEH for rainfall from 1961 - 1990, is virtually the same and can also be used.
• SPR is Standard Percentage Runoff coefficient for the SOIL category.

The ReFH 2 tool provides flow rates and volumes for greenfield and developed sites. The UKSuDS tool for storage estimation uses the ReFH 2 peak flow rates for greenfield runoff along with a few other parameter values. The storage analysis in the UKSuDS tool is not based on the volumes of runoff that can be predicted by the tool for both the predevelopment and post development states. 

The total site area is the whole area within the “red line” boundary being considered for planning and development. The trend for some users to only use the paved area in the UKSuDS tool is incorrect. Rules on excluding large public open space is discussed in 'How should you calculate the “greenfield runoff rate” to the site drainage design?'.

The runoff model which has been used in the UKSuDS tool uses a fixed runoff coefficient of 100% from paved surfaces and 0% runoff from pervious surfaces. This is reasonably valid for areas which are heavily urbanised. The assumption of 84% and 0% respectively (which is commonly applied by users of MicroDrainage and other design tools) is not particularly conservative for assessing storage requirements for extreme events, though for sizing drainage pipework it is probably not unreasonable. This approach was justified in a paper in the 1990s based on the original runoff model in the Wallingford Procedure which was issued in 1983. This justification is a misuse of the correlation equation which had been developed, and has since which been rendered obsolete based on the fact that the original equation was shown to under-predict runoff for large rainfall events.

The assumptions of 100% and 0% is non-conservative once the urban proportion does not represent the majority of the development area and in situations where large volumes of runoff occur from pervious surfaces.

The sewerage modelling fraternity for water companies (as opposed to the developer community) do not use a fixed percentage runoff approach to modelling. They use what is termed the New UK variable runoff equation or the more recent UKWIR equation. These can also be used for drainage analysis for drainage conveyance and storage systems, but they should be used by those who understand how these models should be applied.

The SPR values for SOIL in the IH124 equation can be used, but in addition SPRHOST values can also be applied. However there are a number of issues to consider.

There is often confusion as to the values of the 5 SPR values for the equivalent SOIL classes. These were originally: 0.15, 0.30, 0.40, 0.45, 0.50. As a result of further work carried out in the 1980s by the Institute of Hydrology these were revised to: 0.10, 0.30, 0.37, 0.45, 0.53 respectively.

The parameter for SPR used in IH124 is very similar to that of SPRHOST from FEH. The advantage of using the value from SPRHOST in the IH124 equation is that there are 29 soil classes in FEH (see IH126) while FSR only has 5 soil classes. This information is also available digitally . Thus although the parameters are not directly equivalent, the much higher resolution and degree of accuracy achieved using digitally based soil data means that the use of the FEH parameter is considered to be appropriate. Report IH126 provides information on soil classifications and maps these onto the HOST classes.

Unfortunately there are two sets of SPRHOST values available; those used by FEH and those related to the “predominant soils” classification. The values are very similar for most soil types, but a few have more significant differences. The analysis summarised in Appendix 5 of SC030219 indicates that the latter values would seem to be more appropriate, but a fundamental comparison against river flow data has not been made.

The SPRHOST values range from 0.02 through to 0.6 and this is a wider range than that of SPR for SOIL which ranges from 0.1 through to 0.53. It is advised that a lower limit of 0.1 is applied to SPR although detailed research on this issue has not been made.

In the case of the FEH equation there are no decisions needed in using the equation. FARL can be ignored assuming that most sites do not have ponds and lakes to influence greenfield runoff. BFIHOST, like SPRHOST, does vary for some soils between the predominant soils and those used in FEH. However these differences are very small and of no consequence estimating site runoff.

The Q1, Q30 and Q100 values are calculated for both the IH124 and FEH methods using FSSR 14 growth curve information based on the hydrological region in which the development takes place.

As the ReFH2 site tool provides an estimate of each of these flow rates, these values are used by the tool and therefore these need to be entered.

The ReFH2 provides estimates of greenfield runoff rates and volumes for both the direct runoff and the total runoff as there is an allowance for a base flow for the site. It is recommended that the storage analysis is based on values for the direct runoff and not the total runoff. However the tool will estimate the storage requirements whatever values are provided.

Criteria on flow rate control for site discharge was originally the only form of regulation of runoff. The flow rate discharge requirements ranged widely, but by the early 1990s flow rates as low as 1l/s/ha were being stipulated. This results in massive attenuation storage requirements for surface water runoff management. As a result HR Wallingford carried out a research project to try to establish what flow rate control was needed to be effective, and secondly to consider whether there was an alternative control strategy which was effective in protecting the receiving environment, but which also minimised drainage costs.

This research (Storage requirements for rainfall runoff from development sites, SR580 (2002)) established that, for flow rate control on its own to be effective, the limiting rate of discharge needed to be less than 3l/s/ha (and therefore 2l/s/ha was adopted). This results in sufficient stored surface water remaining on the site during the peak period of flooding in many receiving rivers to provide a useful degree of protection. As a result of this, rules were developed so that drainage design could take advantage of higher discharge rates if volumes of runoff were incorporated into the control strategy. If volume control was not considered then discharges from sites needed to be controlled to a maximum rate of Qbar, but with a minimum limit of 2l/s/ha to protect against excessively low discharge rates.

With the runoff proportion from impermeable surfaces being between 2 and 10 times greater than soils (depending on soil type) during a significant rainfall event, the control of runoff using peak flow rate is only going to have limited benefits as the extra volume of water released must, by definition, result in higher flood levels as it passes downstream as further developments contribute runoff. Controlling runoff volume without control of flow rate may have a local impact, but the attenuation process as it passes downstream would result in flood levels which would be largely unchanged. It can be seen that in spite of the greater emphasis having been placed on flow rate control for the last 30 years, that it is volume control that is perhaps the more important of the two hydraulic control criteria when looking at catchment scale impact.

The current criterion for volume control is only linked to the use of the 1:100 year 6 hour event. This is a very coarse simple requirement aimed at trying to protect downstream communities from flooding. The reason for 6 hours being specified is that small catchments and their streams are likely to be more sensitive to changes in urbanisation and the shortest critical duration of these catchments is likely to be of the order of 4 to 8 hours. In terms of the practicalities of design, in order to achieve control of only discharging the greenfield runoff volume for the 6 hour event at the 1:100 year greenfield runoff rate a design is needed which results in other duration events having limited volumes being discharged at high greenfield flow rates, and also some control takes place for lesser more frequent events

Although there is an emphasis that the runoff volume discharged from the site should be limited to just the greenfield runoff volume, in practice many sites will need to discharge a greater volume than the greenfield volume as infiltration and other volume reduction mechanisms may not be feasible. In these situations, although the 1:100 year greenfield discharge criterion is still permitted to be used for discharges up to the calculated 6 hour volume, the remaining runoff volume needs to be discharged at very low flow rates, usually less than 2l/s/ha.

The volume of runoff for a site for a 1:100 year return period rainfall event is usually between 2 and 10 times the volume of the greenfield runoff. However for return periods of 1 year or smaller rainfall events, the runoff volume factor ranges from 10 times through to orders of magnitude more as there will be virtually no greenfield runoff in many instances. The Interception criterion is therefore aimed at replicating this aspect of greenfield runoff.

The importance of achieving no runoff into rivers for the majority of small rainfall events is considerable in terms of the morphological effects.  Urban runoff for small events have considerable polluting content (sediments, hydrocarbons, metals etc. as well as having elevated temperatures in summer. Discharges may be taking place into a water course where dilution levels are very low and is already in a relatively stressed condition for the fauna and flora within it. The impact of reducing the runoff to zero or close to zero for the majority of these events is to prevent the water quality in the river getting any worse.

Therefore the Interception criterion on volume of runoff for water quality and morphology at one end of the rainfall spectrum and greenfield runoff for the 1:100 year 6 hour flood protection criterion at the other, ensures drainage engineers are encouraged to try and replicate greenfield site response for the site as much as possible for all hydrological conditions without having to carry out a detailed comparison of actual flow rate and volume response for all rainfall events.

Formulae in both FSR and FEH methods, including the ReFH2 plot scale method can analyse a site which has a component of urban area for predicting flow rates. The validity of these approaches is limited to situations where the urban areas is small (around 15%) of the development site, though this is not an apparent constraint for the REFH2 plot scale method. It should be noted that the parameter URBAN (FSR) is very different to URBext (FEH/ReFH2).

As runoff takes place at different rates and proportions and is a function of the size of the event, it might be appropriate to model the two components of the site separately to arrive at an appropriate site discharge rate. The two values derived should not be simply added to give a peak flow rate, but hydrograph methods should be used to show the combined effects as this will often result in a double peaked prediction of runoff flow rate.

There are two methods of approach suggested. The first is to reproduce the previous drainage system  in a simulation model and determine the flow rates discharged from the site for specific return periods (usually 1:1 year, 1:30 year (for sewerage systems) and the 1:100 year event). Constraints are applied such as not including surface water runoff flooding and ensuring the drainage models do not use a pressure head above ground level. Use of the outfall section of pipework and taking the pipe gradient of the final length of pipework as being equal to the pipe gradient is an alternative, possibly conservative, but less accurate approach.

Where details of the site drainage system and the properties it served have been lost or cannot be determined sufficiently accurately, then various assumptions might be permitted in the construction of the model. However because there would be so much uncertainty regarding the actual previous site drainage performance a greenfield analysis could be carried out using high runoff soil type to produce a conservative position. Alternatively the paved proportion of the site could be used by assuming 100% runoff for a 30mm/hr rainfall event. This would give very high rate of flow and should really only be used where the downstream system can take a high rate of flow.

Previously developed sites are sometimes termed brownfield sites. However brownfield is also used to denote polluted or contaminated sites. The term “previously developed” is now generally being used to refer to non-greenfield sites and the different hydraulic criteria applied for site discharge control.  A site which has previously been developed with a drainage system serving buildings or other constructed elements would be categorised as previously developed. In addition sites where buildings which have been removed, but where the site has not been reinstated to greenfield conditions, may be considered to be in this category, but there may be time limits associated with this.
The current position on redevelopment of previously developed sites is that surface water runoff management should provide “betterment” (assuming there was little or no runoff control from the site); usually of the order of 30% is expected, though more than this or even the greenfield rate is always an aspiration. Although this does not often result in a significant discharge constraint (and therefore small storage volumes), it should be noted that government planning guidance does look to reduce flood risk. In some instances this will point to the need to impose greater constraints than meeting a limited “betterment” target.

Betterment is achieved by demonstrating that flow rates discharged from the site are less than those that took place previously on the site. This does mean that this has to be demonstrated usually by using a network model built to reproduce the network performance of the previously existing drainage system, or an approximation of it.

However the exception to this is that Sewerage Undertakers, if their drainage systems are in receipt of the runoff, may wish to address areas of high flood risk downstream of the site and the previous situation may not be considered to be relevant in determining an appropriate discharge consent rate.

It should be noted that although betterment of the order of 30% has generally been seen as being a reasonable target, there is now general awareness in the Environment Agency and other authorities that, in terms of reducing the flooding impact on receiving watercourses, this degree of discharge constraint is probably largely ineffective in protecting the receiving watercourse.
Whatever degree of betterment is used, it is important to note that compliance to Interception criterion for previously developed sites is equally valid as for greenfield sites. This will achieve significant environmental benefits even if flood risks are not reduced.

As for the management of the discharge rate from previously developed sites, volume control is associated with providing “betterment” (assuming there was little or no runoff control from the site); usually of the order of 30% is expected, though the greenfield criterion is always an aspiration. Betterment is achieved by demonstrating what runoff volumes were discharged from the site prior to redevelopment and showing that a reduction has been made. Where the site impermeable surface area has been increased this is often treated as being greenfield development with greenfield runoff rules applied.

It should be noted that the 1:100 year 6 hour event is normally applied in determining the volume of runoff which can be discharged at the “betterment” rate, and additional runoff discharged at 2l/s/ha. To determine the maximum amount of storage needing to be provided requires events of all likely critical durations to be analysed. This does mean that this has to be demonstrated usually by using a network model built to reproduce the network performance of the previously existing drainage system.

Irrespective of control of runoff for large events, it is important that criteria on Interception compliance for previously developed sites is applied as it is equally valid as for protecting receiving waters.

The concept of infiltration is aimed at trying to prevent any runoff taking place from sites when there are small rainfall events. The aim is to minimise the discharge of polluted runoff from entering streams and rivers, particularly in summer periods when they have low flows and the water is warm; conditions which are already stressful for flora and fauna. The emphasis is on achieving no runoff for small rainfall events which are less than 5mm. A small study was carried out by HR Wallingford on Interception (Supermarket representation for SuDS Guidance, Interception storage analysis (2014) using time series rainfall analysis to show how SuDS with little or no infiltration capability could deliver Interception. The three key points to note are:

• Evapotranspiration in summer is a key mechanism for preventing runoff using soil storage;
• Very low rates of infiltration can be effective in significantly enhancing prevention of runoff;
• An Interception criterion of 5mm of rainfall does not mean that rainfall from every event up to 5mm in depth will be prevented from runoff as antecedent conditions for some events will be particularly wet. A probabilistic approach is therefore needed for defining and assessing compliance.

Detailed time series runoff analysis need not be carried out if this is considered to be too complex, and simple rules of thumb can be used. These are based on showing that losses based on using evapotranspiration and marginal infiltration are greater than average daily rainfall (annual rainfall / 365) using summer conditions. Guidance on this is provided in the SuDS Manual.
Rainwater harvesting is another mechanism by which Interception can be delivered. See “How can Rainwater harvesting be used for surface water management?”

The normal position taken by the regulator is

“We do not normally count the storage provided within water butts or rainfall harvesting measures towards the attenuation storage requirements as there is no guarantee that these devices would be empty at the time that a rainfall event occurs”. 

This is a reasonable position to take for standard rainwater harvesting design as it takes no account of the relationship between the demand and supply of the storage unit. For a complete explanation of designing rainwater harvesting systems for surface water control reference can be made to BS8515 Rainwater harvesting code of practice (2009, rev 2013). For a more in depth explanation the report by HR Wallingford “SR736_Developing Stormwater Management using Rainwater Harvesting” (2012) can be studied.

The principle which allows rainwater systems to be designed to provide surface water control (prevent runoff) is based on demand being greater (on average over a period of time) than the supply to it. There is uncertainty associated with the calculations for both the demand and supply elements and these issues are discussed in the HR Wallingford report.

However this rule (demand / supply ratio) is not seen as needing to be complied with if rainwater harvesting is only being used for meeting Interception requirements. A simple check to see that demand is greater than 5mm of rainfall runoff in 2 or 3 days would suffice, on the basis that rainfall only occurs on average about once every three days.

It should be noted that recent trends towards leakage flows from water butts and rainwater tanks to create storage in the system to attenuate the subsequent runoff does not comply with the principle of Interception even though they may have value in providing some degree of stormwater runoff attenuation. 

Interception storage is required in order to ensure that no run-off passes directly to the river for the majority of rainfall events with depths of 5mm or less. This is aimed at trying to replicate greenfield runoff response when no runoff is likely to take place for most small events. This type of storage is principally aimed at river water quality protection - polluted water is prevented from entering the water course for all small rainfall events. A 5mm rainfall threshold which, if effectively applied, will reduce the number of events with runoff into a receiving water body by around 50% and reduce total runoff volumes from the site by a significantly higher proportion.

A probabilistic approach to compliance is required (for instance 80% compliance to no runoff from the first 5mm of a rainfall event should be achieved during the summer and 50% in winter). The use of continuous rainfall series with detailed simulation models can be used to demonstrate effective interception compliance of a site. However where this is considered to be too difficult it can be avoided by the use of various design rules of thumb - see the SuDS Manual (2015).

When rainfall takes place on greenfield sites there is, for the majority of rainfall events during the year (which are small), no discernible surface water runoff to receiving water bodies. The rainwater normally evapotranspires, and in winter it can also result in groundwater recharge. However, impermeable surfaces generate runoff from virtually all rainfall events, and this can have a negative impact on the morphology and ecology of receiving water bodies. Interception is aimed at trying to replicate greenfield runoff conditions. In most urban locations in the UK more than 50% of all rainfall events are less than 5mm in depth, therefore preventing runoff from these events is very significant in terms of both hydrological and pollution impact reduction on receiving waterbodies.

The total annual pollution load from runoff is closely correlated to the total volume of runoff. Therefore prevention of runoff from the majority of all small rainfall events and also reducing runoff volume from larger events contributes very effectively to reducing the pollution load passed to receiving waterbodies. This is particularly important during the summer months when dilution flows in receiving watercourses is normally low.
The Interception criterion is rarely applied, and this is due to the twin effects of lack of awareness of its benefits in spite of the evidence from research, and the knowledge needed in being able to design for achieving it. In trying to meet the Interception criterion it is necessary to apply it in a probabilistic way. No runoff is not expected to be achieved during particularly wet periods when permeable surfaces and soils are saturated, so compliance requirements are set on a probabilistic basis (i.e. Interception should be delivered for a proportion of all events, either on a seasonal or an annual basis). A suggested  target is that 80% compliance to no runoff from the first 5mm of a rainfall event should be achieved for summer rainfall events, but only 50% in winter.

A high level of Interception provided for some parts of the site is not to be considered as adequate compensation for a low degree of interception provision for other locations. Compliance is required for the whole site, or at least road paved areas,  for it to be considered effective.

Interception mechanisms are based on runoff retention. This can be achieved using rainwater harvesting, or using soil storage and evaporation. Either infiltration or transpiration rates can dispose of the runoff from minor events to enable the next event to be captured. Infiltration rates of soils can be very low and still be effective at providing interception.

This criterion reinforces the importance of using vegetative and soil based SuDS components.

The use of continuous rainfall series with detailed simulation models can be used to demonstrate effective interception compliance of a site. However this can be avoided by the use of various design rules of thumb - see the SuDS Manual (2015).

The concept of treatment storage is to provide a body of water in which dilution and partial treatment (by physical, chemical and biological means) of surface water runoff can take place. This is defined as the volume of water which remains in a pond during dry weather periods between rainfall events. The amount of storage normally provided is the volume of runoff from 15mm of rainfall (UK) or using the formula for Vt required by SEPA for Scotland (see the SuDS manual). The Scottish method is also applied in Northern Ireland. 

Treatment storage could theoretically be reduced where Interception storage is provided, or when the runoff from contributing paved surfaces has been effectively pre-treated. However there are no agreed rules or guidance on this issue currently available. 

The concept of treatment volume has been applied to SuDS units other than ponds. However this is rarely a requirement of design and the basis for its application for structures other than ponds is tenuous.

The 1m separation distance between the base of an infiltration component and the groundwater table is considered a pragmatic design approach for infiltration systems:

• It helps ensure that there is an adequate distance between the maximum likely groundwater level and the base of the infiltration device. Groundwater observations are often limited and predicting extreme levels can be uncertain. If groundwater rises above the base, the storage capacity of the infiltration component will be reduced and the performance of the system will change from that for which it was designed.
• 1m of underlying soils provides protection of the groundwater from contaminants which may find their way into runoff entering the component. Groundwater protection is extremely important as it is a very valuable resource and, once it becomes polluted, is extremely difficult if not impossible, to clean. Urban runoff is usually contaminated to some extent and even roof runoff can be polluted. Contaminants may be relatively predictable, or may be very unpredictable – resulting from an accidental spill or from someone using the surface water system as a means of illegally disposing of chemical waste. 

The maximum area that can be discharged to any individual component will be dependent on:

• The size and design characteristics of the component (including any sediment pretreatment systems); and
• The characteristics of the drained area (e.g. % impermeability, slope and land use).

Both of these issues will influence the rate and volume of runoff entering the component, and the likely pollutant load – in particular sediment load as this can causing components to clog or suffer from reduced hydraulic capacity.
Maximum areas are normally specified to ensure the velocities of runoff entering components are non-erosive, sediment accumulation rates are suitably low, treatment performance is effective, and health and safety risks are low. All these aspects are specified within design criteria or component design methods.

A common error which seems to take place is the assumption that all flows for the 1:100 year event are allowed to be discharged at the Greenfield 1 in 100 discharge rate. Occasionally this is increased even further by increasing the flow rate by the climate change factor used in the site drainage design. The following should be noted:

• The 1:100 year greenfield runoff rate should only occur where the runoff volume for the 1:100 year 6 hour event that is discharged at the 1:100 year greenfield rate is limited to the greenfield runoff volume for that event, and any additional runoff only occurs at a maximum of 2l/s/ha. Where the 1:100 year 6 hour volumetric runoff criterion is not complied with, the expectation is that all events greater than the 1:1 year event up to the 1:100 year event, should be limited to a discharge rate of 2l/s/ha. If Qbar is greater than 2l/s/ha then Qbar can be used for return periods greater than the 1:2 return period.
• All events should be managed with the intention of providing like for like return periods for the greenfield discharge  rate. In practice this means that events of the magnitude of a 1:1 year event should be managed to the 1:1 year greenfield flow rate and events between the 1 year and the 100 year flow rate should be managed to provide approximately proportional increase in discharge rates. (Usually only the 1 year and 100 year discharge rates are the consented values).
• No allowance is provided for a climate change factor on the consented discharge rates.

Rule 2 is the least important to comply with as long as flows up to the critical duration 1:1 year return period event (with climate change) is less than the consented rate, and that the critical duration 1:100 year return period (with climate change) event is less than the consented rate.

In the case of large sites there may be outfalls at more than one location; either in the form of receiving sewers or receiving waters. In these situations the predevelopment or greenfield subcatchments must be considered, and also whether further downstream the flows all pass into the same catchment. In principle the fact that there is more than one outfall should not result in a greater flow rate or volume being discharged from the site compared to what would have been discharged for a single outfall for the whole site.

The Flood Estimation Handbook (FEH) 2013 rainfall data is the industry standard estimation of design rainfall. 

The data can be found on the FEH Web Service at https://fehweb.ceh.ac.uk/. The FEH Web Service is a product of the Centre for Ecology & Hydrology (CEH), and is developed, maintained and made available by its agent Wallingford HydroSolutions (WHS).

The Flood Estimation Handbook (FEH) 2013 rainfall data is the industry standard estimation of design rainfall. 

The data can be found on the FEH Web Service at https://fehweb.ceh.ac.uk/. The FEH Web Service is a product of the Centre for Ecology & Hydrology (CEH), and is developed, maintained and made available by its agent Wallingford HydroSolutions (WHS).

‘Source control’ components are fundamental elements of a SuDS scheme.  The term ‘source control’ describes SuDS components that can deliver one or more of the following:

• capture, store and remove (e.g. rainwater harvesting systems, green roofs)
• capture, store and treat (e.g. pervious pavements, detention basins, bioretention systems)
• Capture, convey and treat (e.g. swales)
• Capture, store and infiltrate (e.g. soakaways, infiltration trenches)

It is preferable to implement such systems at or very close to the source of the runoff but, if sites are constrained, runoff may have to be conveyed a distance downstream. In both scenarios, however, the components should still be able to meet the same design objectives and will still need to consider the full range of quantity, quality, amenity and biodiversity objectives and criteria for the system.

Interception is the prevention of runoff from the site for the majority of small rainfall events.   When rain falls on greenfield sites, it produces no discernible runoff for these events.  In contrast, rain falling on impermeable surfaces causes rapid runoff for almost all events – and this can lead to erosion and ecological deterioration of receiving waterbodies.  In addition, runoff from developed surfaces tends to be contaminated with urban pollutants which are then discharged into the environment with every runoff event.  By designing systems to prevent runoff from small rainfall events, receiving waterbodies are protected and pollutants are trapped in components potentially allowing time for them to degrade and/or storing them for future removal.  

By capturing and storing or conveying runoff in soil or aggregate systems – source control components are allowing runoff from small events to be harvested, infiltrated and/or evapotranspired – thus minimising runoff for most small events (unless they occur during long periods of exceptionally wet weather) and delivering on the Interception criteria.

Source control components can also deliver on a range of other criteria/standards – reducing peak flows and volumes, treating runoff, and providing wide-ranging amenity and biodiversity value to sites.  The extent to which they do this will depend on the design.

Infiltration should be used wherever possible, subject to protection of groundwater against polluted runoff. As all runoff is to some degree contaminated, protection is provided by ensuring that some treatment is provided by SuDS systems and that the base of all infiltration units should be at least 1.0m above the highest expected groundwater level. This is often difficult to establish, so groundwater levels should be measured through at least one winter period and a judgement made as to how much higher it might get in a really wet winter.

Where there are important aquifers, or the runoff is particularly contaminated, there may be a need to prevent the use of infiltration, even where the infiltration rate is low. The SuDS Manual and other guidance should be followed where this situation applies.

The critical duration event for streams and rivers range from around 4 hours for very smallest of catchments, through to 8 to 12 hours for many small rivers, going up to 3 days for the largest of rivers such as the River Thames.

As the critical duration for storage systems for developments are usually in the range of 18 hours to 36 hours, the events which cause major flows in rivers are likely to be the same as those which are critical for the storage design for the site. This means that if the river water level affects the hydraulics of the site storage (backwater effects) this aspect needs to be taken into account. The term used is “dependency”.

Where a return period of 1:100 years has to be provided and events are totally independent then the design of the unit can apply the 1:100 year event without considering the return period of the other element (river or tide level). However where there is a dependency or probability of a joint occurrence, then a relevant return period of the second element needs to be used.

The first action to take if it is felt there is a significant level of dependency is to apply a return period of 1:100 years (if that is the relevant criterion) to both elements (representing total dependency) and see what the implications are on storage or whatever measure is relevant. Where the consequences are not significant in terms of say cost or space, then the results of this analysis can be used. However where the impact is considered to be significant, then a more detailed investigation of the combination of return periods should be made using appropriate joint probability analysis and tools to refine the design.

There are a number of drivers for splitting sites into drainage subcatchments, although the size and layout of the site will dictate the value and feasibility of this approach:

• Managing runoff from smaller areas helps keep flow rates and water depths low – normally more appropriate when managing flows in surface systems.
• It is easier to design effective treatment systems when the flow rates and pollutant loadings are relatively low.
• The treatment provided can be proportionate to the pollutant loadings ie. parts of the site with low pollutant loads do not need to have as much treatment as more polluting parts of the site.
• Accidental spills or other pollution events can be isolated more easily and dealt with effectively without affecting the downstream drainage system.
• It encourages pollution ‘ownership’ e.g. responsibility for performance and maintenance can lie with the property owner.
• Poor component performance or component damage/failure can be isolated more easily and dealt with effectively without impacting on the whole site.