Thank you, Jack.
And thanks for the invitation come on and speak with you about this research,
and acknowledge my co-workers in a lot of this research,
particularly at the University of Ulster, Jurg Arnscheidt and Rachel Cassidy
and Catriona Macintosh and technicians.
Following on from Linda's talk, I am really going to be talking about phosphorus.
And in this research about identifying the signals
and also trying to mitigate those signals in headwater streams.
But not to detract from the point that there are nitrogen issues here
and also faecal pathogen issues,
which is linked back to the Waste Directive ruling in Ireland.
I am often asked, because I do a lot of work with diffuse pollution from agriculture,
I am often asked, well what is the contribution from septic systems, for example.
And this is a study from 2005 by Roger Smith and colleagues at Afby.
And it's a good starting point really, which uses a budget which has stood the test of time.
It's a current European Union estimate for phosphorus production per person,
around about 0.75kg per person per year.
If we use it as a starting point and then say,
what is actually retained within a septic system.
Well the only study that we have to do that at the catchment scale
is one by Roger Smith from 1977,
who estimated that septic systems on average would retain about 42% of that phosphorus.
We are seeing that as an export then, an assumed export of 0.4kg per person per year.
Now that again has stood the test of time in the literature that I have been looking at.
Up to this year in fact we see that
the estimates range from between 0.3kg per person per year up to about 0.75kg
if there's no attenuation.
Direct discharges from septic systems into watercourses
we all assume at 0.75kg per person per year.
I always tend to use a range and hedge my bets.
And I think that's environmental realistic to do that.
If I am asked to give a single number I will decline to do that
and give a number based on a range, which is between about 0.4kg and 0.76kg.
But using the 0.4kg per person per year,
we can see that Roger and his colleagues have estimated rural septic tanks
to be about 7% of the phosphorus loads to inland waters in Northern Ireland.
And agriculture nearly 60%.
And I guess what we see from a diffuse pollution point of view is lots of regulation,
and rightly so, to try to mitigate that phosphorus contribution from agriculture.
And that's through things like the farm Nutrient Management Scheme
or Farm Waste Management Scheme in the Republic
and the Nitrates Directive National Action Programme.
But the septic tank contribution really is a small proportion of the pie, or the pie chart.
And I guess that's why it has not received as much attention as the eutrophic impact,
compared with the other sources, especially agriculture.
But the problem is not the load,
and I think we are becoming increasingly aware of that now,
that in any kind of economic analysis of where you get your best bang for your buck
in terms of mitigation, it's not the load that matters with septic tank systems,
it's the contribution to the timing of impact, especially in headwater streams.
And that really is a concentration issue rather than a load issue.
And it's a concentration in low flows.
We see it more in headwater catchments,
and especially in catchments of low permeability.
And that's been mentioned quite a lot this morning.
That the challenge is in these low permeability catchments.
And that impact I would hypothesise at the beginning of this research
that I am presenting now was influenced by the density of systems
in any headwater catchment, the condition of those systems, what state they are in,
the position, compared to nearby drainage features, and also clustering.
That small piece of pie tends to contribute to these issues
to give us these singular point source discharges
which will give us either direct discharge if they are not operating properly,
or a proportion of that discharge if they are operating properly.
And this you know might not be so clear.
Maybe in the future we will have some kind of scratch and sniff slide for these presentations.
But this is raw sewage coming from the previous pipes.
The septic system isn't working properly. It's not working at all.
And sewage is being directly discharged into this receiving headwater.
And when I am talking about headwaters I am talking about catchments
of about 3-5 square kilometres, drainage features at the outlets of those catchments
of about 1 metre wide and maybe,
and depending on whether they have been drained or not, artificially, 0.5-2m deep.
And the work we have been doing is in the Drummin region of the Borders.
And this is a typical aerial shot of that area.
In terms of agriculture it's not as intensive as elsewhere in the country.
It's got a very particular diffuse P pollution issue,
more manifested by soil hydrology rather than by its intensity.
And through this landscape we have a river, which is about 3-4km long.
And an outlet at this end, which discharges into the Ulster Blackwater,
which goes on into Lough Neagh.
Now just looking at that catchment and its river, we see it's a 3 sq.km. catchment.
And again just harking back to these hypotheses about
what the impacts of these septic systems might have,
we can see that if we map all the septic systems
in that catchment on top of the agricultural source,
there's about 29 systems in 3 sq.km. or a density of almost 10 per square kilometre.
You know that's the density. We can assume the density is going to have an impact.
And we can use a number of different methods to identify that impact.
The first one is quite high tech.
And we do this quite a lot in catchment studies now.
We look at data coming out from these catchments in high resolution,
synchronous with discharge, so we can get an idea of load.
And this is a bank side phosphorus analyser at this outlet.
It measures phosphorus three times per hour, sub-hourly basis.
About 75 data points per day.
And along with discharge we can see that if I pick out,
if I am selective with the data and pick out days over the summer, April to September,
days with 0 rainfall, no rainfall, you can see as we expect,
that the discharge declines with time.
And it goes up again every now and again with a rainfall event and then declines down.
What we are seeing here is the establishment of base flow.
And because we are measuring phosphorus at the same time,
we can see that at the same time total phosphorus on the other axis
tends to go up as the base flow decreases.
A classic loss of dilution. It's not rocket science really.
But the concentrations are very high towards the end of the summer,
around about 300 micrograms per litre.
That's 75 data points per day, that's very high resolution, very high tech.
There seems to be a problem here. And we can look at this in another way.
A very low tech solution is to do a simple stream walk.
If we start off at the outlet, at 0 metres on the x-axis,
and walk upstream and take samples where we think a sample should be taken,
and just look at the clustering along that catchment,
we can see that there seems to be a step-change in total phosphorus concentration
around about 2,500 m upstream from the outlet.
And that clustering is manifest on the chart.
Clustering starts here, it doesn't really get any lower,
and a lower density at the top end of the catchment.
You know there's a step-change in phosphorus concentration with density.
But we can also look…that's on one day.
We can look at this over a season by taking specific points
along that river reach and repeat sampling at low flows.
And you know, the hypothesis being that daily sampling should show a similar pattern.
And of course it does.
We get down to the high clustering and the range of total phosphorus concentrations
increases over that time.
And the middle bar in that box spot is the median concentration
or the summer concentrations if you like.
And that's just an important metric to keep an eye on.
The median concentrations is the middle bar of the box plots.
I should also say that some of finger printing metrics we have put in here
have also shown that same pattern.
If you look at the intensity of tryptophan for instance,
if you look at the intensity of boron, the concentration of boron,
and a couple of the other grey water metrics, they seem to follow the same pattern.
Coming to the conclusion that there is a problem with septic system density here,
and discharges, we commissioned a survey of each individual system.
And looked at the type of system, method of discharge, maintenance and operation,
and scored all those systems up to 100 for the worst systems.
It's an inspection survey if you like,
where we scored each system up to 100 for the worst system.
And we got this kind of relationship for this particular catchment,
where the median concentration, the middle bar on the box plot,
compared to the cumulative septic tank score, divided by catchment area,
gave us that very good relationship for this catchment,
for the median concentration at low flows during the summer.
And we did this in two other catchments, the same kind of study.
But we see that the relationship is quite variable.
Now when we looked at these data in a bit more detail
the one big factor that seemed to explain the variability was stream gradient.
The lower the gradient the more attenuation,
the less likely we were to see this in the outlet.
It's just attenuation, some form of attenuation.
Our mitigation policy there was to put in place state of the art systems.
We cherry picked according to the worst system.
We grant-aided them. It's voluntary uptake.
And we had package aeration plants and polishing filters,
depending on the rules in each jurisdiction.
All we could do was influence the condition, not the position or clustering,
or the density, but the condition.
And some of the data that came from that,
and you can read about this in this paper here,
Macintosh and others in Science of the Total Environment,
in bold there we saw the initial density of the systems in 2005/2006, Tyrone the lowest,
Monaghan the highest, and the number of systems in Monaghan
in a 5 sq.km. catchment are particularly high.
And we managed to replace 4, 11 and 3 systems in each catchment respectively,
through voluntary uptake.
However, between us doing this work and our re-survey, building work carried on.
It was the time of the celtic tiger and so forth.
We had 17 extra systems installed in Monaghan, 6 in Tyrone,
but only 1 extra system in Armagh.
Armagh provided some kind of controlled test case here.
But you can see the density on the bottom of that table went up in two of the catchments.
What we saw between the two years of data, remember these are annual statistics,
in flows less than 0.09 cu.m. per second,
The low flows in each catchment are directly comparable in each period,
and consistent of many hundreds of data points,
because it's with high resolution data capture.
We can see that Tyrone concentrations actually went up over the period.
Monaghan went up slightly. And Armagh came down.
It was the only catchment where it didn't increase the density.
And you can see that overall that graph does show that these average concentrations
in low flows are directly correlated with septic system density
in this type of catchment, in this low permeability catchment.
To summarise all that, this first point here,
is something that now I don't have to say too often.
We know it's not the load that's the issue with septic systems,
it's their ability to be diluted into low flows.
They have to be diluted in the receiving streams for there not to be an ecological impact.
And I have seen catchments where, we can see this kind of density,
but with no impact in the stream, because of effective dilution.
The lower the flow the less dilution, the more of a problem there is.
And that fourth bullet point there,
it does seem to be a problem as septic density increases in surface waters.
And I know from a poster outside that there's no correlation between groundwater impact
and density….or so found in that study.
Our mitigation measures to improve condition, we are saying,
are likely to be overwhelmed by density in high risk areas.
It comes down to, what do you do in such an area where,
when you improve the condition, and you have still got density and clustering,
what do you do in those headwaters?
Now there are a number of high cost and low cost solutions to that.
And I guess it's how we view our headwaters ecologically.
They are not a Water Framework Directive scale of monitoring.
But they are important as refugia and also as dilution areas.
And I am also interested in what happens at larger scales.
Is there a natural dilution will occur further downstream?
Thank you very much.
0:00:00 / 0:00:00
Prof. Phil Jordan (University of Ulster)
Rural point sources: Small portions of pie and P in headwater catchments