project

Quickly measuring after work on network

Expert(s):
Kimberly Learbuch MSc, Mirjam Blokker PhD

  • Start date
    01 Mar 2018
  • End date
    30 Jun 2019
  • Principal
    Bedrijfstakonderzoek
  • collaborating partners
    Brabant Water, Evides, WMD, WML

When water utilities carry out work on their distribution networks, there is a risk that dirt, which might contain pathogens, will get into the pipes. For this reason, the activities are conducted very hygienically and the water quality is checked before the pipe is put back into use.

The water utilities currently respect a waiting period of one day between the completion of the work activities and the check of the water quality. The reason for this waiting period has however never been underpinned experimentally. The results of experimental and model research however show that a quicker sampling is actually better. Moreover, in combination with new, rapid analytical techniques, the time before the pipe is cleared for use is cut from a few days to one day. This means the customer experiences less inconvenience and the water utility can move on to the next work assignment sooner. This project will study, under operational conditions, whether rapid sampling provides an accurate picture of water quality.

Is rapid water-quality checking better?

The Hygiene Code for Drinking Water Storage, Transport and Distribution (PCD 1-4-2016) prescribes how a water utility should operate. It requires a waiting period of 12 to 24 hours between the post-works flushing and the water-quality sampling. This is based on the assumption that collecting a sample right after the flushing operation produces an overly-favourable picture of the water quality (it is presumed that the contamination, which at first is localised in the pipe section, is released slowly following flushing, which increases the probability of a positive sample after 12-24 hours). But this assumption was not demonstrably supported by data from practice. Recent BTO research (BTO 2016.017 and BTO 2016.047), as well as several practical measurements taken by Evides, now show that the detection probability of (faecal) contamination is actually higher with rapid sampling, that is, 1-4 hours following the works. According to the model calculations (BTO 2016.017), the detection probability increases from 25% under the current sampling waiting period, to up to 65% for sampling 1-4 hours following flushing.

Two samples for the time being

The water utilities will be taking two samples, one shortly after (2-4 hours) the completion of works and flushing, and one a day later. By comparing the results in which one (or both) of the samples is positive, one can determine whether the rapid sampling is reliable. Since the works are commonly conducted under very hygienic conditions, positive samples occur only seldom. This is why a long period of sampling is needed to collect sufficient data. It is estimated that a whole year of measurements will be required.

Supplying water faster

A shorter waiting period could increase safety if it is confirmed that the detection probability is in fact greater under rapid sampling. In addition, it would make the quicker completion of the works possible so that customers experience less inconvenience, and, lastly, it would mean that subsequent work assignments can begin sooner. If this research shows that a more rapid sampling is actually possible, the hygiene code will be duly modified. Water utilities will then also be able to adjust their work programmes accordingly.