At our support desk for MicroDrainage, we frequently field questions about the rainfall theories. An age-old question is which rainfall theory to use when designing and simulating drainage networks: FSR or FEH.
For drainage designs to be valid, they have to meet the no surcharge, no flood criteria. Pipes have to be sized adequately and effective storage solutions need to be built. To do that, it is vital to understand and accurately represent the inflows entering a drainage network.
That makes it very important to use the right rainfall methodology for each site that you are working on. Get this wrong and it could spell disaster. Your model could predict no flooding, but the reality might be a completely different story.
Remember different rainfall theories will predict different rainfall levels.
In the UK we are rather lucky to have two established rainfall theories:
1. Flood Studies Report (FSR)
2. Flood Estimation Handbook (FEH)
FEH itself has two iterations: FEH99 and FEH13
The question we often find developers asking us is: just which method should we be using?
The debate between FSR and FEH rainfall is a long, on-going one. There are pros and cons for each, and there is no UK-wide legislation stipulating which one should be used. So, what is deemed acceptable by LLFAs across the UK varies significantly.
Note that the percentage differences between FSR and FEH data will vary across the UK. Sometimes the difference will be small and sometimes it may be more significant. The obvious pro for FEH13 data is that is it based on more recent rainfall data, enhanced statistical analysis and a greater number of rainfall gauges compared to FSR data.
We recommend checking which rainfall theory will give you the worst-case scenario on a site by site basis. You can do this in MicroDrainage as it allows both methodologies to be used. FSR is embedded directly, and FEH data can be loaded in via download from the FEH webservice. Early consultations with your LLFA will help you understand which rainfall theory they require drainage designs to be designed with.
Different Approving Authorities may require you to use a particular methodology. In the absence of guidance from the Approving Authority, you should refer to the references for FSR and FEH and the various limitations of each method to make an engineering judgement on the most suitable methodology; often the rainfall method generating the largest intensities is chosen to ensure a conservative design.
Be aware that the rainfall can vary significantly for different areas, durations, seasons, and return periods, and so it may be considered prudent to check with both rainfall methods.
Here is how the methods compare:
This chart shows the ratio of FSR to FEH data for a 100-year, 60 minute duration storm.
Image reference: Kellagher, R., Preliminary rainfall runoff management for developments, R&D Technical Report W5-074/A/TR1 Revision E, Joint Defra/EA Flood and Coastal Erosion Risk Management R&D Programme, January 2012, where the image was reproduced with the permission of UKWIR.
The 'heat map' of the UK is based on a single return period, single storm duration and single seasonal profile. The relationship between FSR and FEH data is more complex than simply the geographic location, and we can show that the FSR/FEH relationship (for a fixed location) varies with return period (in addition it also varies with seasonal profile and storm duration).
The dark red shows where FEH gives ~ 40% higher rainfall intensity compared to FSR. The dark blue shows FEH giving ~ 30% lower rainfall intensity compared to FSR.
The table below illustrates the difference in peak rainfall intensity predicted between FSR and FEH for Oxford, UK for a 60-minute, summer storm.
Common questions and myths
Can you use FEH data for sub-hourly events?
We asked our neighbours in Wallingford, the FEH team at the Centre for Ecology and Hydrology (CEH) about this. They told us that both FEH13 and FEH99 were based on hourly and daily rain gauge data, they did not focus on sub-hourly events.
CEH said that, however, FEH Vol. 2 states that ‘some extrapolation beyond this range is justified, for example for durations as short as half an hour’ and elsewhere ‘it is unlikely that the processes governing half-hour extreme rainfalls are very different to those governing 1-hour extremes’.
They also stated that FEH99 data taken from the CDROM allowed the results to be extrapolated down to a duration of 30 minutes. This was found to be roughly equivalent to the FSR results. The FEH CDROM used to give a warning if a user tried to estimate rainfall from FEH99 for durations of less than 30 minutes. This is because the DDF model was being used to extrapolate depths below the duration at which data was available for the study. However, despite the warning it still provided 15-minute estimates in the export file.
CEH found that there was a user requirement for sub-hourly data. As FSR included sub hourly events, CEH used the ratios of FSR depths for hourly events to the required sub-hourly event to calculate the FEH13 data for sub-hourly events. For FEH13, CEH extrapolated the model down to 5 minutes. There is no warning message when extracting data for durations of less than an hour. We asked CEH for their advice. They told us that sub-hourly values should be used with caution as is the case for FEH99. While an improvement on the FEH99 estimates below one hour, they said that the sub-hourly durations provided for the FEH13 model are therefore outside of the original FEH13 model development.
Myth: you cannot use FEH data for a small site.
This may have been the case when the only option was to use FEH99 data which was purely catchment based. Since the FEH13 release, the CEH has enabled users to select point-based rainfall data instead of catchment based. Rainfall can vary a lot across a catchment.
Myth: FEH data makes models unstable.
We would expect that unstable analysis is more likely to be caused by network layout rather than being attributed directly to FEH data, but this will obviously be model-specific.
Common causes of instability are:
• having very steep pipes (i.e. steeper than 1:10)
• ‘hydraulically’ short pipes (relative to pipe diameter)
• offline loops
• location of storage systems relative to flow controls
• junction type manholes
• overall lack of storage in the network
These are a snapshot of the kind of things that can cause instability, and just because you have some or all of the above present in your model, it does not necessarily mean your model will be unstable. Our best practice advice is to always check the flow, volume and depth vs time graphs for all pipes to check these look sensible and, if in doubt, customers with maintenance can always send their models through to us.
In summary, there is no clear-cut answer on whether to use FSR or FEH, however it seems to be a trend that there is a shift for more use of FEH13 data over FSR. In fact, Scottish Water and many London councils will now only accept FEH13 data. This is possibly due to the fact that it is based on more recent rainfall data, and organisations such as HR Wallingford clearly support the use of FEH over FSR, as you can see in this link: hrwallingford.co.uk/ukstormwaterdrainage
The good news for MicroDrainage users is that both options can be used within the software so you can either ask the LLFA or approving authority for their preferred methodology, or check your designs using both and opt for the worst-case scenario.