Meaningless Procedures

One of the most disappointing features of the review process for the DGR is the way that both OPG and the government invoke empty procedures as reassurance that potentially harmful effects will be avoided. Many of these procedures are invoked without any evaluation or sometimes without any apparent idea behind them – as if simply saying that there is a procedure in place is a guarantee that the procedure will have the effect that it is supposed to.

Nowhere is this more evident than in the consideration of environmental effects. One simple example occurred during the testimony before the JRP on September 26, 2013. One of the environmental regulations about construction in a site like this is that there must be fences surrounding the site to keep reptiles and amphibians from entering it and being killed. In an earlier session, a participant had asked the representative from Environment Canada how well these fences worked at excluding these animals. The representative did not know, so the questions was deferred to a later time. Here is that answer, from the representative from the Ontario Ministry of Natural Resources:

THE CHAIRPERSON: Great. So now that you’re here, the Panel would like to return to a couple of questions that we had originally addressed to Environment Canada.

One of them is a question around, in your experience – in your Ministry’s experience, how well do fences perform in excluding species of concern such as turtles and snakes?

MR. SHOREMAN: I’m not sure… we have performance data, but we do have a policy related to drift fences for reptiles and amphibians. And it was my understanding… I’m fairly certain, that Environment Canada used referenced that policy in applying mitigation measures for amphibians.
Testimony before JRP, September 26, 2013, p. 75-6

So apparently neither the national nor the provincial environmental ministries have the slightest idea whether these fences actually work, nor do they have any data whatsoever on whether they work. However, the provincial ministry does have a policy about the fences, and as long as that policy is being followed they’re happy.

Peter Duinker’s was asked by the JRP to examine OPG’s process for determining cumulative ecological impacts of the proposed project. He found that their process was founded on representing these impacts by means of abstract categories and arbitrarily defined numerical boundaries, to the point where the determinations no longer had any relationship to actual ecology:

As soon as one engages in assignment of ordinal categories to various attributes of a predicted effect, one enters the shifting sands of arbitrariness in professional judgements about limits and combinations.

[For example,] On what grounds has 5% been set as a lower limit for assignment of Low Magnitude? If Medium Magnitude is assigned limits of 11% on the low side and 25% on the high side, what is it about conifer woodlands that would see a Low Magnitude for a 10% reduction, and a Medium Magnitude for anything from 11% to 25%? … There is absolutely no ecological justification for the assignment of these limits, or indeed any such limits. If there were ecological justifications, such as conservation of gene pools of cedar or habitat provisions for important vertebrate species, then surely these would have been provided.
Duinker report, p. 5-6

Instead of grounding its decisions in ecology, OPG prefers to generate a bunch of abstract categories for each VEC and run them through a decision tree. Or rather, for only one VEC (cedar), because OPG decided beforehand that only cedar, out of all thirteen terrestrial VECs, could suffer a residual adverse effect from having a huge construction project proceed in its habitat. Duinker comments:

Thirteen terrestrial VECs are accounted for in section 7 of the EIS. All are predicted to incur measurable changes, but only one is assessed to suffer a residual adverse effect – cedar. A massive and elaborate decision tree is advanced in Figure 7.4.3-1 to show how one might arrive at a conclusion on effect significance. The point of the tree, which includes 19 unique pathways through the maze of considerations of magnitude, extent, irreversibility, timing and duration (curiously, frequency is missing), consequence, and social/ecological importance, is unclear when only one VEC is to be subjected to it. Cedar passes through the maze with magnitude = medium, extent = all, irreversibility = medium/high, timing and duration = low, consequence = low, and finally significance = not significant.

I do not challenge an EIS finding that the residual adverse effect on cedar is not significant. What I challenge is the rationale for making that finding. On what grounds does one judge that consequence is low on the basis that magnitude = medium, extent = all, irreversibility = medium/high, and timing and duration = low? To choose one of the other 18 pathways through the maze, on what grounds would one arrive at a conclusion of “may not be significant” (a term itself not defined) when the assessment says that magnitude = high, extent = medium, irreversibility = high, consequence = high, and social/ecological importance = low? Moreover, I had understood from the Consolidated Responses document (page 471) that all VECs were important – otherwise they would not be VECs and therefore not included in the assessment.

To me, the decision trees represent an unjustified, arbitrary set of pathways to conclusions about effect significance. They are unnecessary for someone to develop a reasoned set of arguments, grounded in an effect prediction and contextual analysis, leading to a conclusion about effect significance.
Duinker report, p. 6-7

Duinker’s conclusions, framed in the terms he was asked to address by the JRP, are:

The EA I examined from the perspective of determination of significance of residual adverse effects has significant flaws of approach and method. Against the criteria I was instructed to use, I find that the analysis embodied in the EIS and Consolidated Responses is:
– not credible – the work does not adhere to what I consider to be a robust approach to determination of significance of residual adverse effects;
– not defensible – the methods include huge elements of arbitrary and indefensible professional judgements;
– unclear – the scientific basis for many professional judgements in setting category limits and decision-tree combinations was not described;
– not reliable – other expert assessors could easily come to different conclusions;
– inappropriate – the methods, as shown above, are unnecessarily complicated and prone to challenge regarding limits and combinations.
Duinker report, p. 7-8

In other words, it is a procedure without meaning. It uses numbers and categorization in terms of importance to sound reliable and scientific; but it has no relationship to actual ecology.

In a similar vein, John Bredehoeft’s presentation examines the efforts to model northern Ontario’s hydrogeology in studies commissioned by NWMO and in the literature. The main NWMO study is by Sykes, Normani and Yin (2011). Their initial attempts to reconstruct the observed water pressure layers in the rock for the proposed site failed, even after multiple attempts using different parameters:

The plot in Figure 7 is typical of the plots of simulated versus observed head at the Bruce Site, obtained by the Sykes group. Using the single phase FRAC3D model the Sykes modeling group could not reconstruct observed heads for the site. The investigators were careful in their modeling and applications, but Figure 7 is a typical result—the match to the observations is bad; the simulated head does not match the observed head. Figure 8 summarizes all of Sykes et al (2011) attempts to match observed heads at the Bruce Site with FRAC3D:
The Sykes group, using their single phase model FRAC3D could not match the observed head at the site.
John Bredehoeft presentation, p. 12-13

However, the program FRAC3D has its limitations, so the Sykes group tried a different piece of software that also models gas flow (the rock in this formation often exhibits 80% water saturation and 20% saturation by gas). By doing so, they succeeded in modeling the observed water pressure – but at the cost of completely misrepresenting the gas:

Sykes et al (2011) used TOUGH2 to simulate two phase water and gas flow at the Bruce Site. In their scenarios it is necessary to introduce gas in the Ordovician rocks…. In one scenario gas was introduced into the Ordovician rocks for a period of 200,000 years, and then stopped. Figure 13 is a plot of simulated head and pressure at approximately 1 million years after the injection ceased. This simulation gets a good fit for the water head data. However when one looks at the simulated gas saturation values, they do not fit the observed data; there is almost no gas left in the rocks.
John Bredehoeft presentation, p. 19

Another major hydrogeological study quoted by OPG is Nasir, Fall, Nguyen, and Evgin (2011). Of this study, Bredehoeft notes

Nasir et al (2011) describe the results:
”First, past glaciation, particularly the second cycle (22,000 abp) had a great impact on pore water pressure gradient and effective stress distribution. The results are consistent with field observations of persistent pressure to the present time. However, the predicted values of anomalous water pressure is less than the observed values at the site, which could be attributed to additional sources, such as gas or somatic pressure.”

The FRACT3D model used by Sykes et al (2011) has similar hydro-mechanical coupling to that used by Nasir et al, but Sykes does not simulate similar under pressures for a glacial loading/unloading scenario. The Sykes et al models and the Nasi et al models appear to show conflicting results, even though their basic hydro-mechanical coupling is presumably the same – the same basic equations are solved.
John Bredehoeft presentation, p. 15-16

So the Nasir group is somehow able to get its model – running on much the same mathematical parameters as the Sykes group’s FRACT3D model – to get what it considers a good fit to the observed water pressures; but it also acknowledges some problems that might be fixed through accounting for gas (which it does not attempt).

Beyond these discrepancies, we would also point out that (as the Nasir quotation mentions in passing) the two groups come to diametrically opposed conclusions on the impact of glaciation on the observed water pressures. The teams were asked explicitly to consider glaciation – not only for its importance in understanding the history of these rock formations, but also in order to assess the possible impact of glaciation in the future.

The Sykes group found no impact:

Nine different paleohydrogeologic scenarios were investigated in this study. Based on these analyses it is concluded that glaciation and deglaciation is unable to yield the abnormal pressures observed in the DGR boreholes. Contributing to this conclusion is the strength of the rock, particularly the Ordovician sediments, and the resulting low storage coefficients that result. From the paleohydrogeologic analyses it is also concluded that flow in the more permeable units such as the Cambrian and Niagaran at the location of the proposed DGR is relatively insensitive to glaciation and deglaciation.
Sykes, Normani, and Yin. “OPG’s Deep Geological Repository for Low and Intermediate Level Waste — Hydrogeologic Modelling” (March, 2011) NWMO DGR-TR-2011-16. p. 237

This led them to conclude very strongly that

Under-pressures were measured in the Ordovician sediments at the DGR boreholes. There is high confidence that these under-pressures are not caused by glaciation and deglaciation.
Sykes, Normani, and Yin. “OPG’s Deep Geological Repository for Low and Intermediate Level Waste — Hydrogeologic Modelling” (March, 2011) NWMO DGR-TR-2011-16. p. 245

Meanwhile, the Nasir group declares

The last glacial cycle in the Northern Hemisphere started approximately 110,000 year ago. During that cycle, southern Ontario was buried under a continental ice sheet, with a maximum thickness of up to 3000 m at about 20,000 years ago. The ice cap retreated approximately 10,000 years ago. However, field data from deep boreholes in sedimentary rocks of southern Ontario show anomalous pore water pressure including underpressure and overpressure zones. In this paper, a large-scale coupled hydro-mechanical (HM) model is developed to investigate the hydro-mechanical (HM) response of the sedimentary rocks of southern Ontario to past glacial cycles. Particular emphasis has been placed on the evolution of pore water pressures and surface displacements. The HM model is verified using analytical solutions. The results of the large-scale HM modeling study shows that the past glaciation, particularly the second cycle (22,000 apb) had significant impact on the pore water pressure gradient and effective stress distribution in the sedimentary rocks of southern Ontario. Furthermore, good agreement between the large scale modeling results and anomalous pressures leads us to the conclusion that these anomalies could be glacially induced. The results of this research can provide information that will contribute to a better understanding of the impact of future glaciations on the long term performance of DRGs in sedimentary rocks.
Nasir O, Fall M, Nguyen TS, Evgin E (2011) “Modelling of the hydro-mechanical response of sedimentary rocks of southern Ontario to past glaciations”, Engineering Geology 123 (2011), abstract

Our point here is not to argue with any of the specific results presented in these hydrogeological papers. The point is that these are scientific investigations that do not match the observed reality and which disagree with each other. And they are simply presented as support for the DGR project without any discussion (or even acknowledgement) of these problems. In citing its hydrogeological research, OPG and the CNSC seem content to be able to point to the voluminous papers it has commissioned, without caring that these papers illustrate the extremely speculative and inexact nature of the field much more than they add up to a case in favor of the DGR proposal. The CNSC actually cites these studies in its Response to Undertaking #18 (Provide additional available research data related to the modelling of anomalous pressure):

In order to provide plausible interpretations of the anomalous pressures at the proposed DGR site, both OPG (Sykes et al., 2011) and the CNSC (Nasir et al., 2011; Nasir et al., 2013; Khader and Novakowski, 2013-under submission) have conducted and supported independent research.
CNSC, Response to Undertaking #18, p. 2

And it goes on to quote small bits of each report without ever indicating that there is any disagreement between them. Once again, CNSC cares much more that a gesture has been made and a procedure followed than whether it actually means anything about the real world. The additional research has been conducted! Whether or not it says anything valid about the DGR project is of no concern.

Comments are closed