Testing of wastewater for COVID-19 has become a routine method for identifying spread of the virus. I was pondering on the challenges of sampling the very large amounts of wastewater that are produced in New Zealand – 1.5B litres per day which is managed through around 320 wastewater treatment plants. From this we need manageable samples that are representative of the wastewater coming out of people’s houses. In Christchurch, for example, there would be around 11.4 million litres of wastewater (I figured out on a pro rata basis across the population as I can’t find accurate figures on line). From this wastewater a sample of 1 litre is collected that is sent to ESR for analysis!
I first came across sampling of this type in relation to coal shipment sampling. Coal shipments being sent offshore are assessed for quality before they leave; if the shipment doesn’t meet the quality standards when it is tested on arrival, no payment is made and the coal is not returned – that’s a costly mistake to make when you are sending 60,000 tonnes of coking coal which could be worth $2m. The way such shipments are assessed is through analysis of around 1kg of material, representing the 60,000 tonnes. A critical part of the analysis is microscopic analysis of two discs of coal slightly largely than 50c pieces. The sampling method for the coal was to collect buckets off the conveyor loading the ship at regular intervals. These buckets were split in an automatic separator into fractions which should be identical, the fractions were combined, then split again and so on, to create the 1kg sample. That 1kg sample was then split further to separate out 50g to put into the analytical samples.
In the same vein, wastewater is sampled by automatic pumps in the wastewater stream, that remove a sample every 15 minutes over 24 hours. These are standard in most wastewater treatment plants, however, if a plant doesn’t have an automated pump a bucket sample is collected through a hatch – not a job anyone would particularly want. The samples are combined. I can’t find a description as to how much is collected, so I don’t know whether the combined sample is homogenised and then a representative portion removed, but in the end ESR receive 1 litre for each wastewater treatment plant.
ESR separate the sample into two and store 500ml so that they can go back and check their analyses if any results are unexpected. They recover virus from both the solid and the liquid waste in the sample. Then they concentrate the material down to about 2.5g by removing water without removing virus. This is 2.5g of sample representing millions of litres of material! After that, they use a standard PCR (polymerase chain reaction) process to analyse for virus – they turn the virus RNA (single stranded, ribonucleic acid) into DNA (double stranded, deoxyribonucleic acid). Nucleic acid is made up of long sequences of only four molecules – cytosine, guanine, adenine and thymine (in DNA) or uracil (in RNA). The order of these molecules determines what chemical actions are controlled by the RNA or DNA. In DNA, the two strands are related in their patterns and linked together – where there is adenine in one strand there is thymine in the other strand or vice versa, where there is cytosine in one strand there is guanine in the other strand. So single stranded RNA can be made into double stranded DNA by finding the right molecular pair and stringing the molecules together.
Once DNA has been constructed from the RNA, it is multiplied up using PCR. This was totally novel when I was a student in the 1980s, now it is routine and can be done in a machine the size of a large scanner on a desktop, or even in a handheld device. The number of times the PCR process is carried out is counted, because the count tells you how concentrated the original RNA/DNA was (in this case the virus RNA). To check their process, ESR add a cat coronavirus and a mouse norovirus to every sample in a known quantity – they measure these viruses as well as SARS-CoV-2 virus (which causes COVID-19) so they can make sure their system is performing accurately.
There are two important factors in a PCR test – sensitivity and specificity. Specificity tells you how certain it is that if the analysis says there is SARS-CoV-2 it is that virus rather than some other virus. PCR is a very sensitive test, except when a virus is at low levels, so ESR is pretty confident that if their analysis says there is SARS-CoV-2, there is indeed COVID-19 in the community. Sensitivity is how sensitive the test is in terms of detecting very low levels of something. Based on their observations to date, ESR consider they can discover COVID-19 if there are approximately 10 cases in an area of 100,000 people. This isn’t a hard and fast rule – some people shed more than others. However, this number is a little disturbing – Christchurch has a population of around 380,000 people. So if a wastewater sample tests positive it could indicate a minimum of around 38 people with COVID-19 who are shedding virus. That’s a lot more than the four who have been identified so far (I don’t know how many people in Christchurch MIQ have COVID-19, but I do think they should be put on porta-potties so as not to confound the sampling).
The uncertainty around single samples and their degree of accuracy is why, if SARS-CoV-2 is detected in an area, the sampling regime moves from a single bulk sample to samples being taken at pumping stations across the region so that a picture can be drawn up of which suburbs have SARS-CoV-2 in their water and at what levels.
Another part of the wastewater picture for me, doing the research for this article, is realising the amount of wastewater needing to be treated every day across New Zealand and the scale of infrastructure required. This is relevant to the current Three Waters reforms – the three waters being wastewater, stormwater and drinking water. 1.5B litres per day and 320 plants treating wastewater are a lot of infrastructure to maintain and keep current. Not to mention that, as populations expand, we need to create new infrastructure, both in terms of all the pipes from people’s houses and businesses and the treatment plants themselves. Christchurch is probably relatively lucky in regard to its infrastructure as the Canterbury earthquake damage meant much of the three waters systems had to be renewed. In contrast, Wellington’s systems are notably failing, perhaps Wellington needs an earthquake to get its Three Waters infrastructure replaced by insurance? I imagine that is not a popular option.
The other reason I wanted to refer to the 3 Waters reforms, is that they seem to be a very evident example of our government going well beyond the mandate our communities would prefer it to have. Every day it seems we hear about a new reform the government is undertaking (yesterday being a new Ministry for Disabled People). It seems like the government thinks it has the ability to reform many aspects of society simultaneously, and without community buy-in, in the case of Three Waters. No-one doubts that we have failing infrastructure and drinking water below standard in numerous areas; however, many doubt that the reforms, with central government strong-arming local government (the majority of local government is opposed to the proposed reform methods), is not the best way to achieve the desired results.
If anything, what we need right now are community-based approaches. Strong-arm, centralised tactics may feel like the necessary approach is times of uncertainty, however that doesn’t mean they are best. One might be put in mind of Trump, and how scared people in the USA voted for a leader who appeared ‘strong’. Look how that ended up…with the leader incentivising the people to storm the place where government meets. With the sentiment in New Zealand at present, our government might be glad the Beehive is in Wellington, where Aucklanders cut off by a border couldn’t storm it if they tried.