The DWMP blog – Episode 12. Are we planning for imaginary flooding?

If you haven’t already seen the earlier episodes in this series, they are all here (https://tinyurl.com/MartinOsborneArticles) I suggest that you start from Episode 1 (https://tinyurl.com/DWMP-blog).

DWMPs are long term plans for drainage and wastewater assets looking at least 25 years ahead.?This requires that they need to take climate change into account.?For flood risk the key impact of climate change is increases in rainfall intensity, particularly in short duration events.?Rise in sea levels will also have an impact in coastal and estuary locations.

A key metric for the DWMPs is the number of properties at risk of flooding for 1:50 annual probability rainfall events.?When modelling the future increase in rainfall intensity due to climate change, the predicted flood risk increases substantially.?Investment plans are being developed to mitigate that future flood risk.?But is that risk real or imaginary?

Why will flood volumes increase?

Flooding is the water that cannot enter the drainage system because it has reached its capacity.?If in the current day the system can take 95% of current day rainfall without flooding then the flood volume is 5% of the total.?If the intense rainfall increases by 10% or 20% in the future, the system capacity doesn’t change and so most of the extra rainfall becomes floodwater, as shown in the diagram below.?

So, the model results are correct that future flood volumes will be much greater than current.

Whose flooding is it?

There is a legal distinction between floodwater coming out of the drainage system because of insufficient capacity and floodwater that cannot get into the drainage system.?From a water company point of view, the former is their flooding and the latter is not.?Water companies would therefore like DWMPs to focus on flooding from the drainage system.?Current models do not represent the limiting inlet capacity of the systems, so they tend to show more of the flooding as being out of the sewerage system than is the case in reality.?Some water companies have considered adjusting the models to limit the inflow to represent the real-world situation.?This would be valid from their point of view, but it does require separate consideration of the impact of water that cannot get into the system and can cause surface water flooding.

Assessing the impact of flooding

More important than the volume of flooding is the impact that it has on properties and infrastructure.?How do we use models to predict this impact?

The impact of flooding depends on exactly how it flows over the ground surface.?It is possible to use a 2D ground model to calculate the route and give a reasonably accurate estimate of where it goes and how many properties it affects.?This approach is time consuming and expensive and requires accurate ground level data.

A simplified approach is recommended in the guidance for the wastewater resilience metric that accompanies the DWMP guidance.?This assumes that the floodwater spreads out in a circular patch from the manhole.?The distance that it spreads depends on the volume of flood water.?

In effect this assumes that the floodwater fills up a conical volume above the manhole.

The number of properties predicted to flood depends on:

• The assumed slope of the base of the cone.?The flatter the slope, the wider the water will spread and the more properties will be affected.?The guidance uses a very flat slope so showing many properties as being flooded.
• The depth of water that is required to flood a property. ?Normal practice is to assume that 150 mm depth is required to get into a property but the guidance says 1 mm is enough.?Again, this exaggerates how many properties are affected.

Although this approach overestimates the number of affected properties both now and in the future, it is only when we consider the future with much greater flood volumes that the predicted number of flooded properties becomes extreme.

An improved approach

I tested the sensitivity of the predicted number of flooded properties to the slope and threshold depth of the flood cone and compared the results with known flooding on a few catchments.?This showed that we should use a cone with a ground slope of 1:40 with a threshold depth of 150 mm for internal property flooding and a threshold depth of 1?mm for flooding outside of the property.?This gives a more reasonable number of properties affected, which seems to match with reported flooding.?It doesn’t indicate exactly which properties will flood as that depends on the exact flow pattern, but it does give approximately the right number.

With this minor change in approach, the future predictions of flood risk are reasonable, not imaginary, and investment should be planned to mitigate them.