The DWMP blog – 32. Does anyone remember critical sewers?

This episode of the blog was triggered by consideration of the Ofwat guidance for the submissions for the PR24 five yearly review of water company business plans (https://www.ofwat.gov.uk/regulated-companies/price-review/2024-price-review/final-methodology/).?This defines 23 performance measures common across all water companies.?These includes the number of “sewer collapses” per thousand kilometres of sewers.?This measure has been in use for some years and I first pointed out its flaws in?a paper to the CIWEM national conference in 2005.

Critical sewers

The first published guidance in the UK on reducing sewer collapses was the Sewerage Rehabilitation Manual (SRM) in 1985 following a spate of major collapses in the 1970s and 80s.

One of the principles of the SRM was that the biggest cost of sewer collapse is the increase in repair cost and the cost of traffic and business disruption.?The flooding and pollution impacts were much less.?

The first step of the SRM approach was to assess the cost, including disruption, of carrying out a repair if a sewer did collapse.?This defined Repair Cost Factors; the ratio of the costs of preventative renovation compared to the post collapse repair.?The driving factors were the depth of the sewer, whether it was in bad ground and whether it was built of brick or was large diameter.?These RCFs were then modified to account for traffic disruption depending on the importance of the road and the number of vehicles using it per day.?There was then an adjustment for sensitive locations including those with a risk of causing pollution, although this was not given a particularly high score.

These overall damage cost factors were then classified into three categories of criticality; values up to 3 were in category C, 3 to 5.9 were category B and above 5.9 were category A.?The theoretical maximum score was 552 so category A covered a very wide range.?Typically 15% of sewers were Category A, 15% Category B and the rest Category C.?Practically every sewer in the country was classified in this way.?

The second step was to inspect category A and B sewers using cctv to identify those that had a significant probability of collapse and then to rehabilitate them to reduce the probability. ?Again, probability can be classified into High, Medium and Low with typical collapse rates of 2% a year for High, 0.2% for medium and 0.02% for low.

Multiplying probability and damage together gives the risk that the sewer poses.?We can colour code these from green for low risk to red for high risk.?The figure below shows the typical proportion of sewers that fall into each category.?As you can see, most sewers are low risk.

Numbers of sewers by risk category

So the SRM focussed on getting the best value reduction of risk by targeting big sewers in bad condition under busy roads.

The Ofwat approach

The definition of the Ofwat measure of sewer collapse is given in https://www.ofwat.gov.uk/wp-content/uploads/2018/03/20190327-7.-Sewer-collapses-final-reporting-guidance.pdf.?I paraphrase.

“A reportable sewer collapse is one that has an impact on customers or the environment.?Impact on customers when the sewer is unable to pass forward flow.?Where there is no impact on customers but there is flooding or pollution this is counted as an impact on the environment.

Collapses that are discovered by the company before they are reported by others are not reportable.”

Note that there is no consideration of the impact on the community or local economy due to disruption in this definition.?So a sewer collapse that results only in a road closure, traffic disruption and damage to the local economy potentially doesn’t need to be reported.?This would mean that the biggest part of the cost is ignored.

Unintended consequences

How would a water company respond to the Ofwat measure of sewer collapses?

It would seek to give the largest reduction in the number of sewer collapses at the least cost.?If we multiply the number of sewers in the previous diagram by the probability of collapse we get the expected number of collapses for each category.

Expected number of collapses by risk category

The biggest number of potential collapses is on sewers with a low damage criticality in Box?1.?This will be small diameter sewers, not under roads, in good ground but in bad condition.?These are also therefore the cheapest to repair.?This gives a strong incentive to focus on these sewers even though they would have little impact if they did collapse.?Meanwhile the Category A sewers tend to be ignored as there are relatively few of them.?The result is that we allow the high impact collapses to occur and spend money repairing sewers that may not collapse and would have little impact if they did.

If we now scale the diagram to show the magnitude of risk (that is probability times impact) rather than the number of collapses then we see the results of this strategy.

Expected damage cost by risk category

We have focussed on the relatively low contribution of the sewers in Box?1 rather than the biggest damage in Box?2.?The result is closed roads and disruption to the local economy.

A better way

So ideally, we want an additional performance measure that takes into account the total damage cost of a sewer collapse as represented by the SRM criticality score.?However, these are no longer regularly used and anyway are difficult to explain to non-specialists, such as economists.?I therefore propose that we use something simpler that captures the disruption, which is the main component of the impact.?This could use the traffic significance of the road as used in the “lane rental” schemes implemented by some local authorities to incentivise utility companies to minimise disruption.?These schemes now include disruption to pedestrians and cyclists as well as motorists, so helping to encourage active challenge.?The charge bands for the Transport for London scheme are shown below.?

If Ofwat implemented a similar idea for a performance measure for traffic disruption and set water companies targets for this, it could encourage a more sensible targeting of sewer repairs to prevent collapses and reduce disruption to local economies.