This is a draft review. Final review will follow after MOE’s response.
K.W.F. Howard M.Sc., Ph.D., P.HG., P.Geo., CGeol FGS
University Professor and Groundwater Consultant
32 Cadbury Court,
Attention: August 28, 2014
Joseph F. Castrilli and Ramani Nadarajah
CANADIAN ENVIRONMENTAL LAW ASSOCIATION
130 Spadina Avenue, Suite 301,
ON M5V 2L4
Dear Mr. Castrilli and Ms. Nadarajah,
Re: Herbicide and Pesticide Concerns – Dufferin Paris Pit
I am providing this brief, summary review of herbicide and pesticide concerns at the Dufferin Paris Pit at the request of Nick Greenacre of the Concerned Citizens of Brant (CCOB). My review is focused on the July 2014 report of Conestoga-Rovers and Associates (CRA) entitled “Assessment of Herbicide and Pesticide Concerns Dufferin Paris Pit”. However, I have also consulted and examined several other associated documents as are listed in the appendix to this letter.
The Paris Pit is owned by Dufferin Aggregates (Dufferin) which is a division of Holcim (Canada) Incorporated. It is located approximately 3 kilometres north of Paris, Ontario. In 1974, the Paris Pit was approved for extraction of aggregates (sand and gravel) by the Province of Ontario and the County of Brant (at that time the Township of South Dumfries) under the provincial Pits and Quarries Control Act, 1971. In 1990, the Aggregate Resources Act (ARA) replaced the Pits and Quarries Control Act, and the Site Plans were revised to reflect the necessary changes needed to meet the ARA Provincial Standards at that time. Subsequently the Site Plans were updated further and approved by the Ministry of Natural Resources (MNR) in accordance with currently enforced standards and policies.
Although licensed for aggregate extraction since 1974, the site has remained in agricultural use, primarily corn production, up to the present time. This type of activity has involved the regular application of various agro-chemicals including the soluble herbicide atrazine, an endocrine disruptor that can cause hormone imbalance. In 2012, Dufferin announced its intention to proceed with quarry operations, raising concerns that aggregate washing at the site would threaten groundwater quality by mobilising the agro-chemicals that had likely accumulated beneath the site. These would include atrazine and its metabolites. To investigate these concerns, CRA was asked by the Ministry of the Environment (MOE) to conduct a site investigation. On the basis of their work (report referenced above) CRA concluded “that there is no credible threat to public or private water supply quality from past use of pesticides at the Paris Pit Site”.
I have closely examined the CRA report and am unable to endorse CRA’s findings. I find the study to be so poorly designed and executed that the range and quality of data obtained is entirely inadequate for drawing appropriate DRAFT conclusions. In its work, CRA focused primarily on the subsurface presence of atrazine and glyphosate. These organic chemicals are moderately soluble in water but tend to bind with soils and sediments when conditions are favourable. Atrazine will adsorb to organic matter that is most likely to be found in the shallow topsoil and in silt and clay horizons of the sediment. Glyphosate will also be found in the topsoil and silt and clay horizons but does not tend to bind to organic matter (as seems to be implied in the study). Instead it binds primarily to cations that are adsorbed in large quantities on the negatively charges surface of clay minerals. If CRA were serious about trying to find atrazine and glyphosate in the subsurface they would have focused their test-pit sampling efforts on these fairly discrete horizons. Instead they chose a more random approach, in which 3 “composite” samples were taken from each pit, by mixing samples collected over several horizons to obtain samples that they considered to be representative of upper, intermediate and lower intervals. None of the analyses for the 9 test-pit samples can be regarded as adequate indicators of the presence (or absence) of atrazine and glyphosate, especially given the sampling method and the sensitivity of the analytical techniques employed (see further discussion re: detection limits at the end of this review).
Samples collected from the borehole cores are said to be “representative of the finer fraction” (which is encouraging) but again the samples are “composite” (i.e. mixed) with no obvious signs of scientific rigour in the sampling protocol and no clear indication of the extent to which samples from potentially discrete atrazine- and glyphosate-rich horizons have been diluted by samples collected from more barren horizons. In total, just 15 composite samples were collected from the borehole cores for atrazine analysis and just 22 for glyphosate. Recognising the site has an area of 260 ha, the “handful” of samples collected and analysed during the CRA study hardly constitutes “extensive horizontal and vertical soil testing” as suggested in the report (Page 9). Overall, the dataset (its size, the sampling protocol, and the detection limits adopted) is simply insufficient to address the serious questions raised about the site.
In an effort to alleviate concerns about the potential release of herbicides to water during the sediment washing process, CRA performs various hypothetical calculations in Section 5.4 of its report. These calculations are over-simplistic at best and simply inappropriate at worst. CRA also seems to miss the point as to where and how the greatest threat to groundwater can be expected to occur. The CRA study does nothing to instil any confidence that sediment operations at the site will be safe from a groundwater quality perspective.
I agree with the suggestion made in the introduction to Section 5.4 that the herbicide concentrations in the water will be influenced by a chemical partitioning process. However, the linear equation provided (essentially S = K x C where C is the concentration of the chemical in solution, S is the concentration of the chemical sorbed to the sediment and K is a partitioning coefficient) is only reliably applicable to very dilute solutions. For most organic chemicals, the assumption of linearity (i.e. a constant value of K) will tend to breakdown when concentrations increase.
The problem is that real sorption isotherms have no prescribed shape (see figure below). They can be linear (as assumed by CRA), in which case the isotherm is straight and K (the slope the isotherm) is constant. However, sorption isotherms can just as easily be concave, convex, or some complex combination of all these shapes, meaning that the degree of partitioning will vary as a function of organic chemical concentration. Without performing batch tests in the laboratory (something not performed in the CRA study) it is difficult to predict what shape the isotherm may take and how partitioning will proceed.
Sorption is normally modeled by fitting an experimentally derived isotherm to theoretical equations. The two most common relationships exhibited by sorption isotherms are the:
where S, K and C are defined above and Q0 is the maximum sorptive capacity for the surface. The Langmuir relationship is particularly important as it demonstrates that, for some sorptive isotherms, there is a limit to which sorption surfaces can exert a significant control on the concentration of the organic chemical in solution.
It also needs to be recognised that:
♦ the state of chemical equilibrium explicitly represented by sorption isotherms is rarely achieved in the field and
♦ partition coefficients will vary spatially in the subsurface, primarily as a function of clay mineral and organic matter content.
With all these considerations, it becomes quite clear that sorption processes of the type proposed in the CRA report cannot be relied upon to limit dissolution of herbicides in the wash water in the absence of significantly more reliable information on the materials present and the precise nature of the sorptive (and desorptive) reactions that can be expected.
In any case, the real issue of concern is not so much whether repeated washing of sediment using the same water will cause a steady increase in herbicide concentration (which may, or may not occur), but whether conditions could be created that would encourage significant transfer of adsorbed chemicals to the water. This issue is not addressed by CRA. In practice, the washing process will produce, over time, many metric tonnes of fine-grained waste material (including silt, clay and organic material) which will remain on site. Because “clean” sand and gravel has been removed, this waste material will host the organic chemicals of concern in concentrations (S) that are likely to be orders of magnitude higher than would have been observed in the original sediment. Quite clearly, water coming into contact with this waste material (for example, in the sediment settling ponds) will, through the partitioning process described by any of isotherms shown above, have the opportunity to acquire very significant concentrations of herbicide. These enriched solutions, over time, represent a very credible threat to groundwater quality in the region. None of the data provided by CRA in its investigation adequately address this overriding concern.
In fairness, CRA do undertake a series of “conservative” analyses of potential impact (Section 5.4) that involve a mass balance approach and ignore the role of sorption. The first of these analyses considers groundwater flow through the wash pond area over a thirty year period and concludes that concentrations of atrazine will remain significantly below ODWQS (by a safety factor of 11 or more). The serious flaw in this calculation is to use groundwater quality determined under “natural” or “pre-quarry” conditions as a basis for predicting groundwater quality following 30 years of severe sediment disturbance. Clearly this is inappropriate.
In another, perhaps more reasonable, mass balance analysis, CRA estimates the concentration of glyphosate in the wash water if all the aggregate washed from the sediment in the course of one year were to have a concentration of 0.0094 mg/kg (the highest value observed amongst the meagre 33 samples (including 3 duplicates) analysed for glyphosate). They found that the concentration would be 0.14 mg/L which is below, albeit marginally, the IMAC for glyphosate of 0.28 mg/L. What is significant here is that 10 of the samples analysed had method detection limits of 1 mg/kg meaning that all 10 could have glyphosate concentrations 100 times higher than the “maximum” used in the calculation without being detected. The remaining samples had method detection limits of 0.005 mg/kg, a value which is only slightly less than that used in the calculation. Recognising that at a more comprehensive, better designed investigation would very likely have revealed values of glyphosate considerably higher than the 0.0094 mg/kg used in the analysis, the safety margin determined for glyphosate (0.14 mg/L versus 0.28 mg/L) can hardly be considered adequate. If a value of 0.005 mg/kg had been used (the lower method detection limit) the safety margin would also be small (a factor of just 3 compared to the IMAC).
Most significant is the fact that a comparable analysis was not carried for atrazine. When a similar calculation is made using a soil atrazine level of 0.009 mg/kg (which is not unreasonably high given that this value lies below the 0.01 mg/kg method detection limit for atrazine), the concentration of atrazine in the wash water is found to be 0.135 mg/L or 135 μg/L. This concentration exceeds the ODWQS of 5 μg/L by a factor of 25! Very clearly, sediment atrazine analyses involving method detection limits of 0.01 mg/kg are completely inadequate for the task being undertaken by CRA.
Another way to consider this very serious problem is that a dry sediment sample with a porosity of 25% and a mass of 2 kg would have a volume of about 1 litre. With a method detection limit of 0.01 mg/kg, the 2 kg sample could contain up to 0.02 mg (20 μg) which is at the very limit of detection. Under conditions of saturation, the 1 litre sample of sediment would contain 250 ml of water. If the atrazine present at the very lowest level of detection were to be transferred to the water, it would have a concentration of 80 μg/L which is 16 times the ODWQS of 5 μg/L. Even a partial transfer due to sorption effects would be significant. Again this demonstrates the inadequacy of the data and the method detection limits adopted.
In summary, CRA was requested by MOE to undertake an investigation that would provide complete reassurance that washing operations would not impact groundwater quality. The study was poorly designed and executed, and has failed to obtain the good quality data so crucially required for reliable interpretation and scientifically justifiable conclusions. Until such time an appropriate, comprehensive investigation is carried out (with adequate detection limits), I believe that, contrary to the assertions made by CRA, there remains a credible threat to public or private water supply quality from past use of pesticides at the Paris Pit Site.
I shall be pleased to expand upon any of my comments or address any other issues should you require it.
Appendix – Documents/Material Consulted and Examined
“Assessment of Herbicide and Pesticide Concerns Dufferin Paris Pit” CRA July 2014 *
Main Report, Figures & Tables.
Appendix A Stratigraphic and Instrumentation Logs
Appendix B Laboratory Analytical Results
Appendix B.1 Tabulated Analytical Data
Appendix B.2 Laboratory Reports
CCOB Docs re assessment:- Background Notes: Water Issues Paris Pit : June 2014 (NG)
CRA Pesticide Assessment Report: Some comments and questions for CCOB (NG August 2014)
WHPA & IPZ map : (generated from GRCA grims.)
Water System Fact Sheet-Paris. GRCA & Brant County
Public Health Ontario: June 2013 : Atrazine & Paris Pit : Report by PH Ontario for Medical Officer of Health.
Abstract : Jablonowski et al 2012 Dry-wet cycles increase pesticide Residue Release from Soil : Env.Tox. & Chem. Vol 31 No. 9 1941-1947, 2012
Abstract : Jablonowski et al 2012 Release of aged 14C-atrazine residues from soil facilitated by dry-wet cycles. Geophysical Research Abstracts Vol 14, EGU2012-9514-1 2012
Concerns Regarding Potential Risks from Agrochemical Residues and Aggregate Washing at Watts Pond Pit with references. (NG Rev. June 2013)
Power Presentation Slides : Concerns Regarding Potential Risks from Agrochemical Residues and Aggregate Washing at Watts Pond Pit (NG & A Ehrlich May 2013) (Presented by CCOB to County Water Engineer, MOE, MOH)
CCOB Water Concerns & Road Costs Flyer….(Summary of concerns and action to end 2013)
PTTW Application http://www.dufferinparispit.com/en/announcements/resources/ParisPit-PermitToTakeWaterApplication.pdf