108 Going (further) down the (highway) drain

After the last episode of the blog on highway drainage I had comments from Rob Cunningham with some insights into why highway drainage did not show up as a major problem in the Reaons for Not Achieving Good (RNAG) data for watercourses in England.  He raised two issues:

1. Scale – the sampling points are generally on large watercourses and do not look at the many small tributaries further upstream.  Polluting discharges into the tributaries may not show up at the downstream sampling points.  This would particularly be a problem for pollutants that readily broke down in the watercourse – such as those in sewage – which would decay before they reached the sampling point.  The pollutants in highway drainage (heavy metals and hydrocarbons) are generally persistent and do not readily break down so this mechanism has less of an effect.  However, the pollutants could be trapped in sediment before reaching the sampling point or diluted by other, cleaner, inflows.
2. Coverage - if a location is not tested for a particular pollutant then the watercourse quality is classed as “Good” for that pollutant, without knowing what its condition really is.

I had not been aware of this second point but it looked like an interesting thing to investigate.

The RNAGs database does not include details of which watercourses were tested for which pollutants but the companion Status Classification data set does.  This includes data for not only the most recent (Cycle 3) 6 year assessment period but also the previous two cycles.  I haven’t given a link to the data as it is huge and is a challenge to analyse.

Pollutants of concern

The RNAGs data seems to show that most heavy metal pollution is related to mining or industry so I have concentrated on the Polycyclic Aromatic Hydrocarbon pollutants.  These seem to the ones of most relevance to highway drainage.  The data has quality classifications for the individual PAHs – Benzene, Benzo(a)pyrene, Benzo(b)fluoranthene, Benzo(g-h-i)perylene, Benzo(k)fluoranthene and Fluoranthene.  (Strictly speaking benzene is not a PAH but is a related pollutant.)  However I have combined these and have classed a watercourse as failing on PAHs if any one of these is shown as a fail.

Results for individual cycles

I have excluded Groundwater and Coastal waters so restricting the analysis to all inland surface waters including lakes and canals.  The table below shows the results for each of the three assessment cycles.  It shows the number of watercourses classified for PAHs, the number where this is based on testing data and the number of those tested that fail.

So what does this tell us?

• The number of watercourses being classified for PAHs showed a big increase from Cycle 1 to Cycle 2.
• In Cycle 3 just under 6% of watercourses were actually tested for PAHs, the rest were assumed to be Good.
• In Cycle 1 and Cycle 3 every watercourse that was actually tested was classified as failing the water quality standard for PAHs.  In Cycle 2 about a quarter of those tested were classed as Good.

I assume that the sites to be tested are selected based on some initial assessment of risk; so we cannot necessarily assume that all 4587 watercourses would fail if they were tested, but it is certainly a concern that most watercourses are not tested for the key pollutants in highway runoff.

Changes from Cycle 2 to Cycle 3

I looked at the changes in classification of each waterbody from Cycle 2 to Cycle 3.  (I ignored Cycle 1 as only 68 waterbodies had actual results.)

The results are very interesting. 

• The reduction in the number of water bodies failing between Cycle 2 and Cycle 3 is because 48 waterbodies that failed in Cycle 2 were not tested in Cycle 3 and so were classed as Good.  However there is no evidence that they had actually improved as there is no testing data from Cycle 3.
• All of the other water bodies that Failed had failures on the same number of PAH components for Cycle 2 and Cycle 3.

Conclusions

We are not seeing the impact of highway drainage discharges in the official water quality classifications because we are not looking for it.  In Cycle 3, only 6% of waterbodies were tested for PAHs that are key pollutants in highway drainage.

The number of waterbodies that are tested for PAHs has reduced from Cycle 2 to Cycle 3 including 48 waterbodies that had been shown to fail the water quality standard and so needed to be monitored.

It would be useful to understand how the sites that are tested and that fail relate to highway discharges.  Can someone with better GIS skills than me please look at this.