104 Reasons to (not) be cheerful – Part 3

A few weeks ago, the UK Environment Agency published the third cycle of Reasons for Not Achieving Good (RNAG) results for waterbodies in England.? This is a requirement under the Water Framework Regulations that require River Basin Management Plans (RBMPs) to be prepared for all water bodies and updated every 6 years.? The RNAG assessment should trigger an action plan of measures to allow the waterbody to achieve Good in the future.

Strangely, I haven’t seen much reaction to the publication either here on LinkedIn or in the press.? Thanks to South West Water for bringing it to my attention.

The last RNAG database used data from 2015 but was published later than that.? The current database does not give a year against the data, but we can perhaps assume that it was from a couple of years ago.

A rant on quality assurance and usability

There has been criticism of water companies and the EA for delaying publication of monitoring data to carry out quality assurance.? Environmental campaigners have a suspicion that the delay is to allow data showing poor performance to be removed to give a sanitised picture.

The EA appears to have taken this criticism to heart and dispensed with any QA on the RNAG data.? It is, frankly, a mess.

Each waterbody or section of a waterbody is assessed against a wide range of measures ranging from the flow regime, through concentrations of pollutants to ecological measures such as healthy fish populations.

My understanding is that each measure should be rated from Bad to High, with a rating of Good being one step below High.? It therefore follows that to be classified as a “Reason for Not Achieving Good” the measure should be less than Good.? In hundreds of cases in the published data it isn’t, with measures that are Good or High being classified as an RNAG.

A new classification has been introduced for this set of data; Reason for Deterioration (RFD).? My interpretation is that this should be measures that are Good, so do not count as an RNAG, but show a deterioration from High.? If so then measures should only be classified RFD if their current status is Good.? Again, there are hundreds that are not.

There are also some classifications that I do not understand.? “Supports good” conjures up an image of fish lined up on the river bank cheering as the river gets a good classification.? Similarly what does “Does not support good” mean?? “Upward trend” is meaningless without knowing what the previous classification and the current classification is.

There is also some sloppy work in allocating the source of the problems.? For example a private sewage treatment discharge cannot be a water company problem.?

I could go on.

I was interested in seeing how RNAGs had changed from the Cycle 2 assessment to Cycle 3.? This should be easy, every RNAG has a unique identifier so just match them up between the two sets and see what has changed.? Except, even if it is a measure of the same pollutant from the same discharge into the same waterbody, it has a different ID in the two sets of data.? So there is no way of matching them up.

Another problem to comparing the data sets is that some data fields have changed their name between the two sets but contain the same type of data and some fields have kept the same (or similar name) but contain a different type of data.

These are all basic checks that should be done as a matter of routine.? So zero out of ten for QA and usability.? Rant over.? Let’s look at the results.

The results

I have looked at two aspects to summarise the results.

• How many waterbodies fail to achieve Good for each pollutant measure.? I have limited this to parameters affecting more than 100 waterbodies.
• How many waterbodies fail to achieve Good for each source of the problem.

I have restricted this to surface waters (canals, lakes and rivers).? Feel free to go and look at the rest.? Bear in mind that I have fixed some, but not all, of the poor data quality so there is still some margin of error.

Measures

I have added together the number for phosphate and total phosphorus as they are just different ways of measuring the same thing.

Note that the total from adding up the columns is much greater than the total number of waterbodies.? This is because most waterbodies are affected by multiple measures.

The change in the total number of water bodies with RNAGs is the balance of 327 waterbodies that have dropped off the list; I assume because they have improved and 518 that have been added to the list.? This could be because more waterbodies have been assessed or because some waterbodies have got worse.? There is another set of data that I would need to cross reference with to check.

The biggest issue by far is phosphate and this is increasing significantly.? Phosphate comes from a wide range of sources including agriculture, sewage, urban runoff and natural sources.

Sources

The table below shows the sources of all of the issues.? Again adding up the columns gives much more than the total number of water bodies.

The biggest cause is agriculture and rural land management and this is increasing.? Second is the water industry.

The large number of “sector under investigation” is a concern.? Perhaps it would be good stakeholder engagement to have an annual report on how many of those have been resolved.? But it isn’t a legal requirement to report that; so maybe it won’t happen.

I have broken down the water industry discharges into type and show the main ones.? (There are also a lot of water industry discharges defined as “other”.? These seem to be mostly associated with water supply infrastructure.)? Continuous sewage discharges are discharge of treated effluent from treatment works.? Intermittent sewage discharges are from storm overflows.? It comes as no surprise to me, but may to others, that intermittent, untreated, discharges are a small proportion of the problems.?

Conclusions

There is a lot of useful information in this RNAGs data; but it does take some digging into.? Maybe it would be easier if I was looking at a particular waterbody rather than national trends.

It would also be interesting to dig further into the “Urban and transport” causes to understand what were the underlying problems.? Maybe a future blog.