Now that you have had a chance to review the input of your fellow panelists (see link in the introduction for a .pdf of results), and given the information about how results from degradation tests should be applied as half-lives in modeling WWTPs, please revisit your confidence ratings for the following: a) What is your confidence in the Rolsky and Kelkar model predictions for polyvinyl alcohol degradation? (1-10) b) What is your confidence in the Rolsky and Kelkar model predictions for polyvinyl alcohol in post-treatment wastewater? (1-10)

a) 5
b) 4

Again, it’s not ideal to ask to give a numerical score to questions like this, as it ignores the nuances of the issue. Yes, there is uncertainty, but that in itself does not detract from the findings – if anything, the uncertainty should lead to a precautionary approach. Q4 above implies that the source of funding might influence the findings of the paper, yet I find that some questions in this review process seem to be pushing for a particular conclusion (i.e. that PVA emissions are less than predicted by Rolsky and Kelkar), when to my mind there is no clear evidence for a shift in the best estimate either way.

It is tempting to adjust my initial score to an even lower level of confidence, but in the expert's assessment I have also seen some arguments in support of Rolsky & Kelskar, or at the very least with a higher confidence score than the scores I awarded. My score was 0.4 and 0.2 below average, respectively. While I may partially disagree with some of the arguments raised by my fellow experts, it would not be scientifically sound to focus only on the arguments that support my own views. Furthermore, the application of WWTP half-lives based on 301-type tests is a huge simplification of reality. While such simplification is required and generally accepted for ERA, it does not mean it is the absolute truth. Exceptions to the half-live conversion will exist, and it is actually reasonable to expect that polymers are more likely to be exceptions to the rule. Since PVA has some sorbtive properties, which could be exacerbated from interaction with e.g., fatty acids, there is room to argue PVA would have a higher half-live than normally used for readily biodegradable materials. After careful consideration, I would score a) the model predictions for degradation a 3 (down one point) and b) the model predictions for PVA in effluent a 2 (down two points). This is primarily due to the fact that the degradation in several stages of the model is likely to be much higher than Rolsky & Kelkar claim, which means the effect on PVA in effluent is magnified. In my initial assessment I was somewhat reluctant to score very strict, but seeing that my views are supported by fellow experts I am more confident to score more strictly.

a (1)
b (1)

a) What is your confidence in the Rolsky and Kelkar model predictions for polyvinyl alcohol degradation? (1-10)

My confidence = 2

b) What is your confidence in the Rolsky and Kelkar model predictions for polyvinyl alcohol in post-treatment wastewater? (1-10)

My confidence = 2

a) 3
b) 3

My confidence was low with two times a score of 2. Actually, my confidence has not changed in the sense that there is some minor uncertainty and this is why I am not inclined to assign a score of '1'.
I think that the degradation in the WWTP and in post-treatment wastewater is underestimated.

a) (2) Confidence was already quite low, such PVA emissions could not be concluded with the lack of data on actual WWTP emissions. The removal data available in the literature is of little relevance. As mentioned above, the data from Menzies et al demonstrate the biodegradability of PVA. Moreover, in Rolsky and Kelkar work, the removal data under aerobic conditions were based on anaerobic operation, which is not a correct assumption.
b) (2) Estimating the effect of post-treatment on PVA removal without using data from specific studies performed with PVA is very risky. Specific experiments are necessary. For example, the authors acknowledge that there are no experiments on the effect of sand filters or percolators on PVA degradation.

a) What is your confidence in the Rolsky and Kelkar model predictions for polyvinyl alcohol degradation? (1-10). My rating: 3

b) What is your confidence in the Rolsky and Kelkar model predictions for polyvinyl alcohol in post-treatment wastewater? (1-10). My rating: 3