Water Quality Data for Western Corridor

The first interim water quality data for the Queensland Western Corridor Recycled Water Project (WCRWP) were released this month.

The full report and an evaluation of it by the Expert Advisory Panel are available from the website of the Queensland Water Commission.

The interim report presents final water quality data (presumably after the advanced oxidation and lime stabilisation) from the advanced water treatment plant at Bundamba. These cover the three-month period of plant validation and verification from 22 May to 25 August 2008 and all subsequent results during normal operation up to December 2008 (an additional four months).


I recommend that interested people download and read the actual interim report. But for those who may care, my own personal comments and interpretation are as follows:

Scope of the water quality analysis

On the range of chemical and microbial contaminants monitored, I think the report is very comprehensive and by far exceeds any water quality monitoring program that I have seen previously from anywhere in the world.

I’m not sure whether the intention is to sustain this level of monitoring. But if that can be done, it will provide a valuable source of knowledge to improve risk assessment and planning for many future advanced water treatment processes.

My only real disappointment with the way that the data are reported is the failure to include analytical detection limits for measurements that were below the analytical detection limit. To report data simply as ‘not detected’ (ND) provides very little information unless the detection limit is known to the reader. From the way that the data are presented, it is clear in all cases that ‘ND’ means that the concentration was less than the Public Health Regulation Standard, but it would be helpful to have some indication of how much less it may have been.

While I understand that this is a summary report (and it is intended to be highly readable to a wide audience), it would also be helpful to have some additional statistical description of the water quality parameters. In some cases, means and standard deviations may have been determinable (where there were sufficient data). In others, the data may have been well presentable as a cumulative probability distribution or other similar means of description.

It would also be helpful to know the concentrations of the chemical constituents earlier in the treatment process. This would allow an assessment to be made regarding the treatment performance of the various treatment ‘barriers’. Such information is useful to give an indication that the individual treatment processes are doing what they are expected to do and thus to properly validate the ‘multiple barrier’ concept.

Physical characteristics

It is a little difficult to interpret the physical parameters without an explicit description of the water sampling location. However, I presume that the water has undergone final stabilisation, which involves the addition of ‘hardness’ (calcium ions) and alkalinity (bicarbonate ions). This explains the relatively high pH (7.5 – 7.8) and total dissolved solids (110 – 170 mg/L) that would both otherwise be expected to be lower directly after reverse osmosis and advanced oxidation treatment.

The stabilisation process is important since ultrapure water is quite ‘aggressive’ and leads to corrosion of pipes. This has the effect of picking up other (less desirable) dissolved substances along the way.

Inorganic compounds

Sixteen cations and four anions were monitored. Of these, the heavy metal cadmium, was observed to exceed the Public Health Regulation Standard on one occasion at the beginning of the validation process. A concentration of 0.0023 mg/L was recorded, compared to a standard of 0.002 mg/L. The explanation given in the report is that cadmium is found is small quantities in the lime that is used for stabilisation and that imprecise lime dosing led to the presence of the dissolved cadmium. This has now been corrected and no such exceedence was again identified following the validation phase or during the operational phase.

Disinfection byproducts

The advanced water treatment process includes a number of disinfection steps, which normally (almost certainly) lead to some formation of disinfection byproducts. The key disinfection processes that may lead to byproducts include chloramination prior to microfiltration, UV/H2O2 advanced oxidation and final chlorination.

The monitoring program included three types of disinfection byproducts. These were inorganic byproducts (bromate and chlorite), organic byproducts (trihalomethanes and haloacetic acids), and nitrosamines (NDMA and NDEA).

One exceedance of bromate (0.04 mg/L compared to standard of 0.02 mg/L) was reported and this was explained to have been the result of a short-term excessive chlorine dose. Corrective action was undertaken on the chlorine dosing system and no further incidences of bromate exceedance have been observed.

Three exceedances for bromodichloromethane reflect the difficulties in regularly complying with the relatively stringent standard of 6 ug/L. There are many drinking water supplies throughout the world that would rarely comply with this standard. From what I understand from the interim report, this standard was introduced in July 2008, after the validation period for the advanced water treatment plant had commenced. As a result, plant operations were adjusted in order to comply with the standard and since then, all subsequent results have been in compliance.

While the NDMA results are technically not an exceedance of Public Health Regulation Standard, the maximum concentration was reported to be 10 ng/L, which is precisely equal to the Standard. In this case, it would certainly be helpful to have a more detailed description of the NDMA concentration distribution. For example, was it commonly within 10% of the Standard, or was this a single aberrant outlier? This is important to help understand the likelihood of exceeding the Standard during future operations (as well as for understanding overall long-term exposure).

Hormones

Eight well-known steroidal hormones were monitored. These included four estrogens (17-alpha-ethynylestradiol, 17-beta-estradiol, estriol and estrone), three androgens (androsterone, etiocholanolone and testosterone) and one progestin (norgestrel).

Many people will be more interested in the results for hormones than I am. In my opinion, the fact that none of the hormones could be detected was inevitable given the source water quality and the nature of the advanced treatment processes.

The issue of risks associated with hormones in advanced water recycling schemes has been severely exaggerated by certain politicians who apparently don’t mind looking foolish for the sake of whipping up a little hysteria.

Nonetheless, the results for the hormones provide a useful illustration of my earlier comment regarding detection limits. The Public Health Regulation Standard for 17-alpha-ethinylestradiol is given as 1.5 ng/L. Depending on the analytical method used, it is quite likely that this value is very close to detection limit. Thus it would be helpful to have an indication of how far below the standard we can be confident of being.

Other organic chemicals

None of the five chemicals presented in this category were reported to have exceeded the Public Health Regulation Standard on any occasion. Those chemicals for which actual numbers (as opposed to ‘ND’) were reported indicate that a very significant gap exists between the standard and actual measured concentrations.

Microbiological water quality

Results for Escheria coli and clostridium perfringens spores indicate excellent disinfection of bacteria across the multiple barrier system. The more difficult micro-organisms to manage are viruses. The non-detection of somatic coliphages is an indication of good control of viruses, but the two exceedances for F-RNA phages do raise the eyebrows.

The interim report describes how these results were further investigated and states:

“On balance, it is concluded that these two detections were the result of the analytical method used and did not indicate the presence of bacteriophages in the purified recycled water.

In line with the findings of the investigations, the following corrective actions have been implemented:

• duplicate samples are being taken, to provide greater certainty of results;
• the number and frequency of water quality tests will be increased where an initial positive result is returned; and
• changes have been made to the sampling and analysis process.

“False” positives will continue to be reported even if duplicate analyses return a negative result. Continuous monitoring, plant shut down controls, and additional water quality testing are part of the ongoing plant operation and form part of the robust risk identification and mitigation practices.

I’d suggest keeping a close eye on this parameter in order to improve our understanding of both the analytical reliability and the nature of its removal/inactivation by the various individual barriers of the advanced treatment processes.

Additional PCR testing for a range of specific viruses (rotavirus, astrovirus, noroviruses, adenovirus, enteroviruses, hepatitis A and reovirus) provides some assurance of the absence of these highly pathogenic organisms.

Herbicides, pesticides and phenols

Thirty five herbicides, pesticides and phenols were monitored, but none of then exceeded the relevant Public Health Regulation Standard. Without reported detection limits, it is difficult to interpret how low expected concentrations may be for most of these contaminants.

I would be interested to compare this result with current water quality in Lake Wivenhoe…

Pharmaceuticals and personal care products

Fifty five pharmaceuticals and personal care products were monitored, but again, none of them exceeded the relevant Public Health Regulation Standard.

Some of these results would be particularly useful to relate to concentrations prior to individual treatment processes. Some, such as caffeine and salicylic acid can provide a very good indication of reverse osmosis treatment performance and thus it is useful to monitor their removal (even if they are well below Public Health Regulation Standards).

Radionuclides

No exceedance of radioactivity was observed.


QWC Expert Advisory Panel comments

The Interim Water Quality Report is preceded by a letter of assessment by the QWC Expert Advisory Panel. The letter states that:

“Based on the testing results in the report, the commissioning of the Western Corridor Recycled Water Project is proceeding well, providing confidence it is capable of consistently producing purified recycled water that is safe to be used to supplement Wivenhoe Dam.

The results indicate that the treatment process barriers are effective in controlling water quality hazards and reliably producing purified recycled water suitable for release into Wivenhoe Dam. No exceedances of the water quality standards have been measured in this testing data after normal operations commenced.”

I agree with this assessment. However, I'd like to see some more raw data, purely for the purpose of trying to draw some further information and insights from it.

I think it is very important not to get carried away with monitoring end-point water quality. In my opinion, much more important information can be gleaned by carefully observing individual treatment processes and ensuring that they are each operating effectively. This is the whole basis of the multiple-barrier treatment philosophy. Without closely monitoring each individual barrier, it is not possible to have confidence that if one barrier fails, then another will provide the necessary redundancy to ensure safety.

I’m not suggesting that suitable multiple-barrier monitoring is not being undertaken, -I presume that it probably is. However, I’d like to raise awareness that this is the type of data that we should really all be interested in, -rather than just long tables of ‘non-detects’… more on that topic soon!