The DWMP blog – Episode 14. How much overflow discharge is too much?

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).

I will now consider the emotive topic of storm overflows.?My intention here is not to justify why we need storm overflows (see my other LinkedIn posts for some thoughts on that) but to consider how we incorporate planning for storm overflows into DWMPs.

Storm overflows are important features of combined sewerage systems (that is those that take both wastewater and rainwater) to prevent flooding of properties and other areas and to protect the treatment process from being washed out by high flows.

Storm overflows can potentially cause several types of impact to the water environment.

1.???Aesthetic impact from solid material such as wet wipes.

2.???Bacteria and pathogen causing a health risk for activities involving contact with the water or consumption of shellfish.

3.???Impact on the ecology and wildlife.

The first impact can largely be managed by screening the flow before spill to remove large solids.?The second is managed by reducing the frequency of spill or ensuring that the spilled flow is dispersed away from water contact activities.?The third is avoided by ensuring that the flow is sufficiently diluted with rainwater before it is spilled that it does not have an unacceptable impact on the environment (typically diluted to roughly the same concentrations as treated effluent).

However, storm overflows are also perceived to be “a bad thing” and there is pressure to set an arbitrary limit on the number of times that they spill each year, irrespective of any requirements to meet the targets set out above.

So, how do we consider these types of management strategies in preparing a DWMP?

Screen design

It is not practical to build a screen of a size able to take the absolute maximum flow that can ever be discharged.?Normal practice is to design for an event with a 1:5 annual probability of occurrence.?This means that every few years some unscreened flow will be discharged.?The design would use 1:5 annual probability design storms to size the screen and a 1:30 event to size the maximum flow.?In theory the maximum flow could be greater in an extreme event, but in practice the additional flow would not get through the sewerage system (see Episode 13).

Spill frequency

Spill frequency is assessed with a long record of representative rainfall; a “timeseries”.?This can either be a record of actual data over several years or a synthetic series generated from a statistical analysis of recorded data.?The latter is able to generate long series (typically 40 years) of represented rainfall from shorter rainfall records or records with gaps.

To predict future conditions the timeseries needs to be modified to represent the influence of climate change and there are methods available to do this; although they are not always robust.

The number of spills will vary from year to year depending on the rainfall, so how many years of timeseries do you need to use to accurately assess the average spill frequency.?The more years you need, the longer and more expensive the analysis and the less opportunity to refine the plan and repeat the analysis.

The answer depends on what the target is for average number of spills per year.?If the target is more than 10 spills per year, then one representative year will represent enough of the pattern of rainfall variation to give a reliable prediction.?If the target is 3 spills per year, which is typical for coastal bathing water standards, then you need to analyse at least 10 years of rainfall to correctly represent the range of conditions.?If the target is 1 spill per year, which has been suggested for river bathing waters, then you need to analyse at least 20 years of rainfall data.

Having decided how many years of data you need, there is then the challenge of picking the appropriate years from the starting point of a 40-year series.?That might be the topic for a future blog.

Ecological impact

The most important chemical parameters affecting river ecology are ammonia, biological oxygen demand (BOD) and dissolved oxygen.?

Ammonia is a naturally occurring pollutant that is present in sewage and can be toxic to fish.?The highest concentration of ammonia will be just downstream of the discharge before it has time to decay.?Ammonia occurs in two forms; highly toxic un-ionised ammonia and less toxic ionised ammonium.?The proportions of the two depend on the temperature and the pH of the water.

Biological oxygen demand is a measure of the amount of oxygen used up over many hours by the decay of naturally occurring chemicals in sewage.?Its measurement normally excludes the oxygen taken up by the decay of ammonia.?The highest concentration of BOD will be just downstream of the discharge before it has time to decay.?

Dissolved oxygen works the opposite way.?The presence of dissolved oxygen is a good thing as it allows fish and other aquatic life to flourish.?Low concentrations damage aquatic life.?Dissolved oxygen is increased by turbulence dragging oxygen from the air to dissolve in the water.?It is reduced by the decay of ammonia and BOD.?The effect of the decay on dissolved oxygen is known as the oxygen sag curve.?The worst case for dissolved oxygen will occur several hours later than the discharge and therefore several kilometres downstream.?

High concentrations of ammonia combined with low dissolved oxygen has more impact than the sum of its parts.?For an individual discharge this is unlikely to be a problem as they occur in different locations but for multiple discharges the low dissolved oxygen from one can overlap with the high ammonia from another further downstream.?

The impact on the ecology is defined by the worst short-term conditions.?There are two approaches used in different situations.?The first for less sensitive locations:

1.???The 99 percentile standards define concentrations that cannot be breached for more than 1% of the year – that is a few days.?These can miss short term impacts that can still affect the ecology.?These standards are set for ammonia and BOD and so are assessed just downstream of the discharge point where concentrations are worst.?These are assessed with a timeseries of flow and other conditions and normally one year is used as this is adequate to represent the few days of exceedance that is permitted.

2.???The Fundamental Intermittent Standards (FIS) define the concentrations that can only be exceeded infrequently for short duration, including for the worst one hour in a year. ?This is a much more rigorous standard as the worst hour is only 0.01% of the year compared to 1% for the percentile standards.?These standards are set for ammonia and dissolved oxygen and so the worst cases for ammonia and dissolved oxygen are at different locations.?These are again assessed with a timeseries but 10 years is normally used to correctly represent the infrequent exceedance.

The Environment Agency consenting guidelines allows assessment against only the percentile standards if the discharge is well diluted by the river.?However, the more recent Storm Overflow Assessment Framework (SOAF) require compliance with both sets of standards in all cases.?I think that this was a mistake and that compliance with only the percentile standards should be required for less sensitive locations.

Summary

Assessing overflow performance is much more time consuming and costly than assessing flooding and this affects the number of potential investment options that can be assessed and how good the resulting plans are.?Simpler methods allow more time for thought and refinement of the plans.?Complex, time consuming methods risk becoming a tick-box exercise rather than real planning.

The next blog will cover the geeky details of how the ecological impact should be modelled.