September 30, 2016

MEMORANDUM

To:          Andrea Donlon, CRWC

David Deen, CRWC

From:    Laura Wildman, P.E., Princeton Hydro, LLC

Paul Woodworth, Fluvial Geomorphologist, Princeton Hydro, LLC

Melinda Daniels, PhD, Fluvial Geomorphologist, Stroud Water Research Center

Re:         FERC Re-Licensing Process for TransCanada Hydro Northeast Inc.

Peer-Review of ILP Study 2 and Study 3

Riverbank Transect and Riverbank Erosion Studies

The Connecticut River Watershed Council (CRWC) is a stakeholder and participant in the re-licensing process of the Federal Energy Regulatory Commission (FERC) for the three hydropower facilities owned by TransCanada Hydro Northeast Inc. on the Connecticut River, Wilder Dam, Bellows Falls Dam, and Vernon Dam.  Princeton Hydro, with the Stroud Water Research Center, was retained by CRWC to complete peer-review of technical erosion studies, specifically Integrated Licensing Process (ILP) Study 2 and Study 3: Riverbank Transect and Riverbank Erosion Studies.  ILP Study 1: Historical Riverbank Position and Erosion Study was reviewed for background data, as was the study plan laid out in the Revised Study Plan (RSP), dated August 14, 2013, and as revised in Appendix B: Staff’s Recommendations on Proposed and Requested Study Modifications And Studies Requested, dated September 13, 2013[1].    This memorandum is a critical review of ILP Study 2 and Study 3 and aims to address the following questions as defined in 18 CFR § 5.15 Conduct of studies (d) Criteria for modification of approved study, and the RSP:

  • Were the studies completed as per the Revised Study Plan?
  • Were the objectives set in the RSP met?
    1. If not, is additional data collection or analysis warranted?
  • Were the methods described in the RSP utilized?
  • Was the analysis described in the RSP conducted?
  • Was the Study conducted in a manner consistent with generally accepted scientific practice?
    1. Was the methodology utilized consistent with generally accepted scientific practice?
      1. If not, is additional analysis or a different type of analysis warranted to meet the RSP goals of conducting the study in a manner consistent with generally accepted scientific practice?
    2. Were the conclusions of the study consistent with the scientific evidence presented?
  • Were the deliverables promised in the RSP included in the final study report submittal?

FRAMEWORK FOR THIS PEER REVIEW

For ease of review of this memorandum we have italicized, placed in quotes, and referenced page numbers for any text taken directly from the Revised Study Plan (RSP) or the combined Study 2 and Study 3 Report.  Our comments have been structured as per the Integrated Licensing Process (ILP) regulations 18 CFR § 5.15(d)(1) regarding conduct of studies, and have been subcategorized to reflect the structure of the subsections taken from the Revised Study Plan, dated August 14, 2013, pages 19-36, and additionally revised September 13, 2013.

The Revised Study Plan was organized into 14 sections, including:

RELEVANT STUDY REQUESTS, STUDY GOALS AND OBJECTIVES, RELEVANT JURISDICTIONAL AGENCY RESOURCE MANAGEMENT GOALS, ASSOCIATION WITH OTHER STUDIES, EXISTING INFORMATION AND NEED FOR ADDITIONAL, INFORMATION PROJECT NEXUS, STUDY AREA AND STUDY SITES, METHODS, ANALYSIS, CONSISTENCY WITH GENERALLY ACCEPTED SCIENTIFIC PRACTICE, DELIVERABLES, SCHEDULE, LEVEL OF EFFORT AND COST, REFERENCES

We have organized our review to comment on only those sections of the Study 2 and Study 3 Report that we felt were not conducted as provided for in the Revised Study Plan, as per the ILP regulations 18 CFR § 5.15(d)(1) regarding conduct of studies.  The sections we commented on relating to their consistency with the RSP are:

STUDY GOALS AND OBJECTIVES

METHODS

ANALYSIS

CONSISTENCY WITH GENERALLY ACCEPTED SCIENTIFIC PRACTICE

DELIVERABLES

All of our comments fall under the ILP regulations 18 CFR § 5.15(d)(1), which asks if “approved studies were not conducted as provided for in the approved study plan.”

Each section of our review starts by including the exact statement from the Study 2 and Study 3 Report that we are commenting on, and then follows with our peer review comment and our recommendation.

STUDY GOALS AND OBJECTIVES

This section includes our comments on the “Study Goals and Objectives” as described in the RSP.  We have specifically commented on the objectives from the RSP that we feel were not met or not conducted as provided for in the RSP.

Objective from RSP, under Study Goals and Objectives: “Observed water-level fluctuations and shear stresses from nonproject-related factors will also be investigated.” (Page 19, RSP Study 2)

“Hydraulic modeling (Study 4) will be integrated into the study after field sampling ends to analyze the relationship between shear stress and bank erosion.” (Page 25, RSP Study 2)

“Analyze hydraulic modeling data to provide information on flow velocity, stage (water surface elevation or WSE), and shear stress impacting riverbanks in the study area.”  (Page 5, Study 2 and Study 3 Report)

Peer Review Comment: No hydraulic modeling results, including shear stress impacting riverbanks in the study area, were analyzed or discussed in the Study Report.  Without this analysis, a key part of the study as proposed in the RSP is missing and a fundamental driver in the erosion process (i.e. shear stress) has gone unassessed.

Recommendation #1: TransCanada should incorporate hydraulic modeling results from Study 4 into Study 2 and Study 3, and analyze the results to assess the relationship between shear stress and river bank erosion, as proposed in the RSP.

Objective from RSP, under Study Goals and Objectives: “The objectives of this study are to: Ascertain the likely causes of erosion (e.g., high flows, groundwater seeps, eddies, and water-level fluctuations related to project operations).” (Page 27, RSP Study 3)

“This study will ascertain the relative importance of water-level fluctuations associated with project operations in the erosion process relative to other contributing factors and how the importance of water-level fluctuations in the erosion process varies with soil type and geomorphic setting.” (Page 21, RSP Study 2)

Peer Review Comment: The third objective of Study 3, to “ascertain the likely causes of erosion” (page 27, RSP Study 3), has not been completed, nor has the study ascertained “the relative importance of water-level fluctuations associated with project operations in the erosion process relative to other contributing factors” (page 21, RSP Study 2).  The Study 2 and Study 3 Report characterizes the cyclical processes of bank erosion but concludes that, “Trying to distinguish specific effects of normal project operations among the panoply of potential controls on bank erosion in any given location is not possible“ (page 108, Study 2 and Study 3), and states, “Attempting to identify a single cause for erosion fails to recognize that multiple processes operate collectively to effect change on the riverbanks through space and time” (ES-3, Study 2 and Study 3 Report).  The fact that there are multiple causes of bank erosion is a generally accepted assertion; the intent of the study was not to determine if project operation were the single cause for erosion but to ascertain the likely causes of erosion, in other words to ascertain which causes are more dominant than others and thus, to “facilitate conclusions as to the association and effect of project operations on active erosion” (page 13, RSP Study 1).  The RSP’s Project Nexus for Study 2, page 21, states that, “This study will ascertain the relative importance of water-level fluctuations associated with project operations in the erosion process relative to other contributing factors”. The study fails to “ascertain the relative importance” of the project operations (i.e., WSE fluctuation) in relation to other contributing factors (e.g. high flows, groundwater seeps, eddies), because it uses a methodology that cannot accomplish this study objective, referred to in the study as “the erosion ratio” (first described on page 82, Study 2 and Study 3 Report), and which was not proposed in the RSP.  Please see our comment under the section on “CONSISTENCY WITH GENERALLY ACCEPTED SCIENTIFIC PRACTICE” in this peer review.

In addition, no data was collected to ascertain groundwater seeps associated with water fluctuation as a likely cause of erosion.  An investigation of groundwater seeps would have required identifying the elevation of groundwater adjacent to the banks with respect to the varying water surface elevation in the channel.  The report states, “the magnitude of water surface fluctuations in the study area is less than 2.0 ft for 75% of the study area’s length so hydraulic gradients between groundwater levels in the bank and the adjacent river level are likely small” (page 111, Executive Summary, Study 2 and Study 3 Report); however, no groundwater data was collected to affirm that statement, nor to assess the remaining 25% of the study area.

The Revised Study Plan provides a simple list of causes of erosion, “e.g., high flows, groundwater seeps, eddies, and water-level fluctuations related to project operations,” (page 27, RSP Study 3) that were considered at the outset of the project.  However, the Study does not consider adjacent land use as a factor (other than the presence or absence of riparian vegetation at the top of bank), and yet numerous peer-reviewed research studies have investigated and confirmed that adjacent land use has a strong role in bank stability and erosion.  In addition, the study does not consider the impact that daily WSE fluctuations may have on limiting vegetative growth at the toe of the river banks, although the study itself acknowledges the important role that vegetation can have on increasing bank resistance to erosion.

Recommendation #2: TransCanada should re-evaluate the existing data, or if necessary gather additional data, with respect to these important factors (i.e., methodology used, groundwater elevations, and surrounding land use) to “ascertain the relative importance of water-level fluctuations associated with project operations in the erosion process relative to other contributing factors” as per the RSP (page 21, RSP Study 2).

Objective from RSP, under Study Goals and Objectives: “Identify the effects of shoreline erosion on other resources (e.g., riparian areas and shoreline wetlands, rare plant and animal populations, water quality, and aquatic and terrestrial wildlife habitat).” (Page 27, RSP Study 3)

Peer Review Comment: The fourth objective of Study 3, “to identify the effects of shoreline erosion on other resources” (page 27, RSP Study 3) has not been completed.  In the final section of the Study 2 and Study 3 Report, the Assessment of Project Effects makes brief references to other studies (page 112, Study 2 and Study 3 Report); however, these studies do not assess shoreline erosion project effects and in most cases these additional studies were not intended to do so.  Specifically:

  1. With regard to Water Quality, reference is made to Study 6 – Water Quality Monitoring Study (Louis Berger Group and Normandeau, 2016a), which “found that the Wilder, Bellows Falls, and Vernon projects had negligible to no effect on turbidity” (page 112, Study 2 and Study 3); however, the following statement, “the few recorded spikes in turbidity were found to occur in response to high flows resulting from heavy rain events,” (page 112, Study 2 and Study 3) fails to distinguish if bank erosion is a contributing factor in the turbidity peaks. Thus, the project effects on water quality remain unassessed and its conclusion that project operations had negligible effect on turbidity are unfounded.
  2. With regard to Aquatic Habitat, reference is made to Study 8 – Channel Morphology and Benthic Habitat Study (Stantec and Normandeau, 2016), but acknowledges that the study did not quantify the effect of fine-grained riverbank materials on increased embeddedness of coarse-grained spawning substrates. Another reference is made to Studies 14/15 – Resident Fish Spawning in Impoundments and Riverine Sections Studies (Normandeau, 2016a), Study 16 – Sea Lamprey Spawning (Normandeau, 2016b), and Study 21 – American Shad Telemetry Study – Vernon (Normandeau, 2016c) (page 112, Study 2 and Study 3); however, none of these studies had the objective of assessing the impacts of bank erosion on aquatic habitats.  Thus, the project effects on aquatic habitat remain unassessed.
  3. With regard to Rare Animal Populations, reference is made to Study 24 – Dwarf Wedgemussel and Co-Occurring Mussel Study (Biodrawversity et al., 2014; 2015, Study 25 – Dragonfly and Damsel Fly Inventory and Assessment (Normandeau, 2016d), Study 26 – Cobblestone and Puritan Tiger Beetle Survey (Normandeau, 2016e), Study 28 – Fowler’s Toad Survey (Normandeau, 2016f), and Study 29 – Northeastern Bulrush Survey (Normandeau, 2016g) and they “did not identify erosion resulting from normal project operations water level fluctuations as a potential factor” (page 113, Study 2 and Study 3). However, none of these studies had the objective of assessing the impacts of bank erosion on rare animal populations, both direct (i.e. WSE fluctuation) and indirect (i.e. bank collapse impacts).  Thus, the project effects on these rare animal populations remain unassessed.

Recommendation #3: TransCanada should revise the Study 2 and Study 3 Report to identify the effects of shoreline erosion on riparian areas and shoreline wetlands, rare plant and animal populations, water quality, and aquatic and terrestrial wildlife habitat, as stated in the RSP.

Objective from RSP revision of Sept. 13, 2013: The study’s analysis will include a correlation of visible indicators of erosion with project-caused water-level fluctuations at the 21 transect locations established in the Riverbank Transect Study (Study 2).” (Page 1 Study 2 and Study 3)

Peer Review Comment:  This objective is not accomplished because the “erosion ratio” metric (page 82, Study 2 and Study 3) employed to attempt to identify correlation is not a generally accepted scientific practice.   It lacks the rigor of other accepted statistical analysis techniques.  For additional discussion on this topic please see our comments relating to “CONSISTENCY WITH GENERALLY ACCEPTED SCIENTIFIC PRACTICE” later in this memorandum.

The RSP notes on page 32 the importance stratigraphy can play in bank erosion: “the layering of sediments within the banks can play an instrumental role in bank stability with contacts between permeable sand above impermeable clay providing a zone along which water can preferentially seep out of the bank.  Consequently, identification of the various sedimentary layers within a bank is critical to understanding the distribution and causes of erosion.” In addition, the RSP states on page 29 that, “Detailed information to be collected as part of this study on bank stratigraphy, depth to sand-clay interfaces, and their relationship to past water-level fluctuations is needed to confirm whether project operations are causing reductions in bank instability.”

While stratigraphic data were collected and provided in the appendices, these data were only referenced in general statements in the study and not analyzed or discussed, such that the relationship between WSE fluctuations and bank instability were unassessed.  The Study 2 and Study 3 Report states:

  • Banks composed of non-cohesive sediments and interlayered cohesive and non-cohesive sediments are the most susceptible to erosion.” (page 10, Study 2 and Study 3 Report)
  • “Normal project operations result in daily or sub-daily fluctuating water levels. At many sites, the position of those daily fluctuations on the bank aligns with the location of notching at the base of the bank(page 53, Study 2 and Study 3 Report).  Figure 5.4.2-6, below, from the Study 2 and Study 3 Report illustrates this observation, with the location where the WSE fluctuation based on normal operating range intersects with the notch in the river bank, circled in red.
  • “Fluctuations in WSE related to normal project operations under no-spill conditions are consistent with notching and overhangs observed at the base of 8 of the 21 monitored banks at some point during the two-year monitoring period (Appendix A). Erosion can result from seepage forces generated by WSE fluctuations (Budhu and Gobin, 1995) with overhangs developing when seepage is focused along a single layer (Fox and Wilson, 2010).” (page 111, Study 2 and Study 3)
  • “The character of sediments in the study area creates banks with limited resistance to erosion. The bank sediments at the monitoring sites, representative of the study area as a whole, are nearly ubiquitously comprised of fine-grained and unconsolidated floodplain or glaciogenic sediments particularly prone to erosion (see Appendix A stratigraphic columns).  Frequently observed inter-beds of permeable sand and less permeable silt can further reduce the resisting force of floodplain sediments by creating horizontal surfaces along which groundwater can preferentially move, potentially increasing seepage forces acting on the bank.” (page 109, Study 2 and Study 3)

A more detailed discussion and analysis of these site conditions is warranted in order to determine “their relationship to past water-level fluctuations” and “confirm whether project operations are causing reductions in bank instability” as per the plan set forth in the RSP.

The study concludes that, “Trying to distinguish specific effects of normal project operations among the panoply of potential controls on bank erosion in any given location is not possible,” (page 108).  We suggest that a statistical method such as an Analysis of Variance (ANOVA), Multivariate Analysis of Variance (MANOVA), or Principal Component Analysis (PCA) would be consistent with generally accepted scientific practice and would yield more conclusive results. This is further discussed in our comments under “CONSISTENCY WITH GENERALLY ACCEPTED SCIENTIFIC PRACTICE” that follow in this memorandum.  In addition to a statistical method, a modeling method, such as Bank Toe Erosion Model (BSTEM), coupled with a sensitivity analysis of the input variables, would assist in distinguishing the degree to which WSE fluctuation impacts bank stability versus other erosive mechanisms.  The input variables in a bank stability model such as BSTEM typically include geotechnical and vegetation data, such as surface erodibility, critical shear stress, geotechnical strength, bulk unit weight, riverbank sediment particle-size distribution, maximum rooting depth of vegetation, and riparian species distribution.  This type of bank stability model has the ability to run with various parameters either included or not included, in order to better assess the likely causation of erosion.

Recommendation #4: TransCanada should utilize the existing data to further assess the potentially “instrumental role” that WSE fluctuation may have on initiating the erosion cycle, by directly comparing the elevations where notching is observed and where the normal operational WSE fluctuations occur, and incorporate their data, relating to the “layering of sediments within the banks” and the stratification of permeable and less permeable zones, into this assessment.  In addition, TransCanada should utilize a more rigorous statistical method to analyze the significant amount of data collected.

METHODS

This section includes our comments on the “Methods” as described in the RSP.  We have only included comments on the sections of the “Methods” from the RSP that we feel were not conducted as provided for in the RSP.

Repeat Surveys

Statement from RSP: “TransCanada will consult with the erosion working group during the 2-year monitoring period to discuss the need for, and locations of, increased sampling frequency based on the initial monitoring results and any information gleaned from the historical data research in Study 1 (Historical Riverbank Position and Erosion) that supports the need for more periodic monitoring based on significant erosion rates.  The need for, and extent of, additional monitoring approaches (e.g., groundwater-level monitoring) could also be discussed in consultation.”  (Page 23 RSP Study 2)

Peer Review Comment:  TransCanada did not consult with the Erosion Working Group[2] during the 2-year monitoring period as described on page 23 of the RSP.  The Erosion Working Group participated in choosing the transect locations, but was allowed only to review the study after the 2-years of monitoring were completed and the Study 2 and Study 3 Report were submitted.

This interim consultation appears to have been added to the RSP to justify the reduction in the number of monitoring sites from 30, requested by FERC (10 for each project), to 20 (page 21 RSP Study 2, an additional cross section was added later) and from a biweekly monitoring frequency, requested by NHDES, NHFG, and VANR, to “at least four times per year for 2 years” (page 23 RSP Study 2), and yet this consultation and interim reporting did not take place.

Recommendation #5:    TransCanada should formally meet with the erosion working group as necessary to consider its comments and revise the Study 2 and Study 3 Report to reflect those comments, as proposed in the RSP.

Hydraulic Modeling

Statement from RSP: “For this study, two-dimensional (2-D) modeling at up to six sites using River2D may be necessary to understand complex sites where HEC-RAS modeling does not adequately describe eddy flows that might develop, for example, upstream of valley constrictions or flow deflection that might occur, for example, around a mid-channel bar or island.” (Page 33 RSP Study 3)

Peer Review Comment: The RSP stated that 2D modeling “may be necessary to understand the complex sites”.  No 2D modeling was prepared, nor was its use or reasons for not using it discussed in the Study.

Recommendation #6:  TransCanada should add a discussion to the Study 2 and Study 3 Report that explains why 2D modeling was not completed and that the 1D modeling provided in Study 4 was adequate to analyze the more complex sites

ANALYSIS

This section includes our comments on the “Analysis” as described in the RSP.  We have only included comments where we felt that the Study Report was not conducted as provided for in the RSP.

Statement from RSP: “TransCanada will consult with the erosion working group periodically to solicit comments to strengthen data collection procedures, analysis of erosion causes, and continuing studies during the 2-year study period.” (Page 35 RSP Study 3)

Peer Review Comment: TransCanada did not consult with the Erosion Working Group periodically to solicit comments to strengthen data collection procedures, analysis of erosion causes, and continuing studies during the 2-year study period for Study 3, as stated in the RSP.

Recommendation #7:  FERC should consider the August 1, 2016 Study 2 and Study 3 Report to be the interim report and that the Erosion Working Group’s current review of Study 2 and Study 3 Report be integrated into a revised study that the Erosion Working Group is then able to review as the final study as proposed in the RSP.

CONSISTENCY WITH GENERALLY ACCEPTED SCIENTIFIC PRACTICE

This section includes our comments on the Study 2 and Study 3 Report in relation to its “Consistency with Generally Accepted Scientific Practice” as described in the RSP.  We have only included comments on the methods used and conclusions drawn that we feel were not conducted as provided for in the RSP. We have broken our comments down into two sections, to respond to the Study’s consistency with generally accepted scientific practice, as stated below:

“The various methods to be used for this study conform to generally accept scientific practice” (Page 24 RSP Study 2), and “The various methods to be used in the Riverbank Erosion Study conform to generally accepted scientific practice as detailed in the Methods section above” (Page 35 RSP Study 3).

The first section relates to comments on how the methodology used in Study 2 and Study 3 is consistent with generally accepted scientific practice (i.e., cross section selection and the erosion ratio method utilized), the second relates to comments on whether the Study’s conclusions are supported by the evidence given.

  1. Consistency of Methodology with Generally Accepted Scientific Practice

Cross Section Selection for Monitoring

While we believe that the study selected the cross section monitoring sites in accordance with the RSP, the study extrapolates observations regarding bank erosion, on a project-wide basis, from monitoring sites that were “selected so a range of” conditions “are incorporated into the analysis” (page 22 RSP Study 2).  Because the sites were not selected to reflect statistical occurrence along the project-wide reach, any extrapolation on a project-wide basis may not be well supported.

Recommendation #8: TransCanada should revise the report to omit extrapolations to the entire study area based on the monitored cross sections unless a statistically based method is used to link cross-section observations with their likely occurrence frequency over the entire study reach.

Erosion Ratio

The primary metric relied upon for “identifying the propensity of erosion to occur in association with certain conditions” (Page 34 RSP Study 3) is the “erosion ratio” (Page 82, Study 2 and Study 3).  This approach is not a generally accepted scientific practice and is not included in the RSP.  It was presumably used to accomplish the objective of including a “correlation of visible indicators of erosion with project-caused water-level fluctuations at the 21 transect locations” (Page 1, Study 2 and Study 3).  No citation or reference is provided for this metric, and the metric is not used, to our knowledge, in the extant fluvial geomorphic scientific literature.  The Study does not demonstrate that the method “conforms to generally accepted scientific practice” (page 24, RSP Study 2 and Page 35 RSP Study 3).

The erosion ratio is too simplistic for attempting to ascertain the likely causes of erosion, when there are multiple known causes.  It is defined as the ratio of two percentages: “the percentage of bank erosion in the study site that is present within a specified feature divided by the percentage of bank length occupied by that feature” (page 82, Study 2 and Study 3).  According to the report, a value greater than 1.0 represents a propensity (or “more likely to occur”), and a value less than 1.0 indicates no propensity (i.e. “less likely to occur”).

Generally accepted scientific practices for analyzing processes with multiple causative variables rely on statistical analyses more sophisticated and robust than simple ratios.  Such statistical methods that may be applied in these Studies, depending on the type and structure of the collected data, include Analysis of Variance (ANOVA), Multivariate Analysis of Variance (MANOVA), or Principal Component Analysis (PCA).

As much of the data is geographic in nature (e.g. the location of bank instability) and managed within a GIS, more rigorous spatial statistical methods should be employed to ascertain spatial autocorrelation or spatial regression, particularly to analyze “correlation of visible indicators of erosion with the project-caused water-level fluctuations” (page 1, Study 2 and Study 3).

For example, the data set that categorizes all banks within the study area into one of six classes: eroding, vegetated eroding, failing armor, stable, healed erosion, and armored (page 79, Section 5.6.4 Mapping Results, Study 2 and Study 3) should be re-analyzed through one of these multi-variate methods with respect to the various contributing factors such as bank height, WSE fluctuation, riparian vegetation, bend geometry, etc. It also should be noted that the classes “healed erosion” and “armored” are essentially banks that were eroding in the past but are not anymore, and those previously eroding banks may also have been due to project operations.  Including those two classes in the larger “stable” category may lead to overlooking past impacts associated with project operations and could significantly change the findings in Section 5.3 Analysis of Historical Aerial Photography (page 23).  For example, in Figures 5.3-1a, b, c, (Pages 25, 27, and 29, Study 2 and Study 3, respectively) it is unclear whether the decrease in bank erosion through time was influenced by bank armoring which may have arrested the process in some areas, while the source of the problem continued to exist.

In addition, the data set derived from the review of aerial photographs at 0.5-mile increments (Page 23, Section 5.3 Analysis of Historical Aerial Photography, Study 2 and Study 3) should also be re-analyzed with multi-variate methods with respect to the bank classification data mentioned above or with respect to the various contributing factors such as bank height, WSE fluctuation, riparian vegetation, bend geometry, etc.  Further, the data set related to the 21 transects (Page 30,  Section 5.4 Erosion Monitoring, Study 2 and Study 3) should be re-analyzed through a multi-variate method with respect to the erosion at the top, upper, mid, lower and toe of bank (Table 5.4.2-1, page 44) and to the median WSE fluctuation.

The erosion ratio appears to have limitations and be subject to biases.  In discussing the erosion ratio associated with WSE fluctuations in the Vernon impoundment in Section 5.6.5 of Study 2 and 3, page 97, an abnormally high value is dismissed because the WSE range in question exists for such short lengths, which indicates that the erosion value can be easily skewed.   To avoid ‘interpreting results potentially skewed by short lengths’, the analysis deliberately disregards any bank lengths that are less than 10% of the study area.  This is problematic for two reasons.  First, it overlooks banks that, albeit short, may be severely impacted by project operations.  Second, multiple classes of 0.5-foot increment WSE fluctuations, which may fall below the arbitrary 10% threshold individually, collectively add up to a significant proportion, likely over 25%.   Thus, in attempting to circumvent allegedly skewed results, the analysis dispenses with data that could otherwise be informative.

Recommendation #9: TransCanada should re-analyze the data in Study 2 and Study 3 Report according to generally accepted scientific practice, as specified in the RSP.   The data set that categorizes all banks within the study area into one of six classes (i.e. eroding, vegetated eroding, failing armor, stable, healed erosion, and armored (Page 79, Section 5.6.4 Mapping Results, Study 2 and Study 3)), should be re-analyzed through a multi-variate statistical method with respect to the various contributing factors such as bank height, WSE fluctuation, riparian vegetation, bend geometry, etc. In addition, the data set derived from the review of aerial photographs at 0.5-mile increments (Page 23, Section 5.3 Analysis of Historical Aerial Photography, Study 2 and Study 3) should also be re-analyzed with multi-variate methods with respect to the bank classification data mentioned above or with respect to the various contributing factors such as bank height, WSE fluctuation, riparian vegetation, bend geometry, etc.  The data set related to the 21 transects should be re-analyzed through a multi-variate method with respect to the erosion on the bank (at the top, upper, mid, lower, and toe) and the median WSE fluctuation.

  1. Consistency of Conclusions with Scientific Evidence Presented

Included below are our peer review comments relating to the consistency of the conclusions stated in the Study 2 and Study 3 Report.  We believe that many of the Studies’ conclusions were accurate and reflected a sound review of the significant amount of data, both historic and current, that was collected.  Our comments below therefore only focus on Study conclusions that are not properly supported by the data presented in the Studies, or were not stated in conjunction with other related findings.    

Study Conclusion #1: “Taken together, natural conditions in the study area, by both reducing the resisting forces and enhancing the driving forces, create a situation where the riverbanks are likely near the threshold of erosion. As a result, minor natural or anthropogenic changes in the study area have the potential to initiate erosion already primed by the river valley’s natural history and character.” (Page 109 Study 2 and Study 3)

“Given the significant changes in the rate and amounts of erosion documented through historical aerial photography and multiple mapping efforts, respectively, normal project operations that have changed little in several decades cannot adequately explain the observed patterns of erosion. Attempting to identify a single cause for erosion fails to recognize that multiple processes operate collectively to effect change on the riverbanks through space and time.” (Page 115 Study 2 and Study 3)

Peer Review Comment: The study points out the significance of river banks that are at the “threshold of failure” by stating on page 11 “When a bank is at the threshold of failure, a slight increase in shear stress or a small decrease in shear strength can lead to bank erosion”.   The study then concludes, on page 109, that the riverbanks in the study area “are likely near the threshold of erosion” and that “As a result, minor natural or anthropogenic changes in the study area have the potential to initiate erosion already primed by the river valley’s natural history and character.”   These statements further support the need to confirm whether project operations are playing any role in the reductions in bank instability.  Because the study area has been classified as being near the threshold of failure, analysis of the data does not support a conclusion that dismisses the significance of the potential role of WES fluctuation in the cycle of erosion based on the fact that it is not the “single cause for erosion” (page 115 Study 2 and Study 3).

Recommendation #10: TransCanada should revise the Study 2 and Study 3 Report to assess how the “threshold” conditions of the study reach may be impacted by even the slightest change in erosive force, whether acting alone, or in conjunction with other erosive forces.  

Study Conclusion #2: “The apparently increasing rate of erosion in the upper Wilder impoundment (Figure 5.3-1a) is more likely related to upstream inflows than Wilder project operations. The upper Wilder impoundment is closer to the McIndoes project than to Wilder dam.  Therefore, McIndoes inflows along with significant natural discharges likely have a greater impact on erosion rates in upper Wilder impoundment than Wilder project operations.” (Page 111, Study 2 and Study 3, Assessment of Project Effects)

Peer Review Comment: This statement is not supported by any data included and described in Study 2 and Study 3.

Recommendation #11: TransCanada should provide their data on the upstream inflows in the Wilder impoundment and analysis to support their conclusion regarding the impact of these inflows.

Study Conclusion #3: The fact that these three sites experienced recession only once during two years of monitoring and that 12 additional monitoring sites mapped as unstable did not experience any bank recession at all may seem incongruous but actually indicates that bank recession, even in the most unstable areas monitored, does not occur annually but rather occurs episodically at time scales extending more than two years.”  (Page 52, Study 2 and Study 3)

“Fluctuations in WSE related to normal project operations … are consistent with notching and overhangs observed at the base of 8 of 21 monitored banks at some point during the monitoring period.” (Page 111, Study 2 and Study 3)

Peer Review Comment: Section 5.4.2 Repeat Monitoring indicates that only three of the 21 monitored transects experienced measurable recession at the top of the bank, and that erosion does not occur annually but rather episodically at time scales beyond the 2-year monitoring period.  The study also acknowledges: “At many sites, the position of those daily fluctuations on the bank aligns with the location of notching at the base of the bank: (page 53 Study 2 and Study 3), “Fluctuations in WSE related to normal project operations … are consistent with notching and overhangs observed at the base of 8 of 21 monitored banks at some point during the monitoring period” (page 111 Study 2 and Study 3).  These observations seem to indicate that as many as 38% (8/21) of monitored banks may experience notching as a result of project-caused WSE fluctuations.  According to the final study the notching at the base of the bank likely drives the “idealized cycle of erosion“ depicted in Figure 5.6.2-1 and results in eventual top of bank recession.  It is noted therefore that this episodic erosion could be related to WSE fluctuation and may not be able to be adequately assessed within a 2-year monitoring period.

Recommendation #12: TransCanada should extend the cross section monitoring beyond the two-year monitoring period proposed in the RSP, for the above reasons and because the Study itself indicates that this period was not long enough to analyze the “cycle of erosion” at all sites.

Study Conclusion #4: “The magnitude of water surface fluctuations in the study area is less than 2.0 ft for 75% of the study area’s length so hydraulic gradients between groundwater levels in the bank and the adjacent river level are likely small, whereas waves breaking against the bank at the same elevation as water level fluctuations may generate stronger erosive forces.” (Page 111 Study 2 and Study 3, Executive Summary and Assessment of Project Effects)

Peer Review Comment: Hydraulic gradients depend on the elevation of surface water and groundwater, which were not measured.  As there was no assessment of hydraulic gradients, wave actions, or erosive forces, no valid comparison can be made between the two bank erosion factors. This statement also raises the question of whether daily WSE fluctuations increases the vertical range on the bank that becomes exposed to wave action and ice jams and their associated erosive forces.

Recommendation #13: TransCanada should retract this conclusion, unless additional data is supplied that supports this statement.    TransCanada should analyze how the WSE fluctuation may increase the vertical range on the bank that is exposed to additional erosive forces such as boat waves, piping and ice jams, which are all issues identified in the RSP.

Study Conclusion #5:The approximately 40% of bank instability mapped through the study area is similar to more free-flowing portions of the Connecticut River (Field, 2005), so normal project operations cannot be considered to be a cause of excessive erosion.” (Page 114, Study 2 and Study 3)

Peer Review Comment: Throughout Section 5.3 Analysis of Historical Aerial Photography (page 23 Study 2 and Study 3) and in subsequent sections, comparisons are made between impounded sections and riverine sections with the assumption that conditions in the riverine sections are natural, normal or unaffected by project operations.  However, riverine sections are also subject to the downstream effects of dams, which includes exacerbated / accelerated bank erosion due to sediment trapping by the dam and sediment deprivation in the downstream reaches.  While these Studies are not focused on the downstream effects of the dams, this “hungry water” effect (Kondolf, 1997) renders any conclusions from such comparisons invalid.

Recommendation #14: TransCanada should provide additional data in Study 2 and Study 3 Report regarding the previous assessment of free-flowing portions of the Connecticut River, if it is to be used as a scientifically supported comparison to the impounded reaches.  Specifically, TransCanada should show how these “free-flowing” reaches are not impacted by other factors such as limited upstream sediment inputs due to the presence of upstream dams.

Study Conclusion #6: Tractive forces generated by flood flows are the only mechanism capable of removing the sediment from the base of the bank that otherwise would lead to bank stabilization if not removed.” (Page 114, Study 2 and Study 3, Conclusions)

Peer Review Comment: This statement is not supported by any data included and described in Study 2 and Study 3.  This statement speaks to the importance of flood flows and tractive forces in the “cycle of erosion” described in the study; however, it appears that no attempt was made to quantify the shear stress created by flood flows or to utilize relevant data from Study 4 – Hydraulic Modeling Study.

Recommendation #15: TransCanada should complete additional analysis of the hydraulic conditions.

Study Conclusion #7: “While other processes such as waves or seepage forces created by project-related WSE fluctuations may exert some control on the cycle of erosion, they cannot be considered as resulting in excessive erosion that negatively impacts other resources since ultimately the continuation of erosion depends on flood flows that sustain the cycle of erosion.” (Page 114, Study 2 and Study 3, Conclusions)

Peer Review Comment: This study conclusion does not follow a logical thought process since although it is stated that “seepage forces created by project-related WSE fluctuations may exert some control on the cycle of erosion” the role of their impact cannot be negated based solely on the fact that these initial erosive forces are taking place in a riverine environment where high flows are ultimately transporting eroded material downstream and continuing the cycle indefinitely.

This statement discounts the role of fluctuating WSE on bank erosion because it is likely acting on only a portion of the “cycle of erosion”; however, the study describes a “cycle of erosion” that is initiated with the creation of a notch or overhang at the toe of the bank (see Figure 5.6.2-1).  Further, the description from the Executive Summary states that “Bank erosion in the study area is a cyclic process that begins with the formation of notches and overhangs at the base of the bank.  The resulting over-steepening at the bank’s base destabilizes the upper bank generating planar slips, rotational slumps, topples, and flows that transfer bank material downslope.  Material supplied from the erosion of the upper bank accumulates at the base of the bank and can ultimately lead to the stabilization of the bank unless the sediment and fallen trees are removed by river currents, wave action, groundwater seepage, or other forces.  If the material is removed, the notching at the base of the bank can begin afresh and the cycle of erosion repeated.”  (Page ES-1, Study 2 and Study 3)

The study also reports that “Erosion can result from seepage forces generated by WSE fluctuations (Budhu and Gobin, 1995) with overhangs developing when seepage is focused along a single layer (Fox and Wilson, 2010)” (page 111 Study 2 and Study 3).

Recommendation #16:  Based on the data presented, TransCanada should revise the statement as follows:

“Processes such as waves or seepage forces created by project-related WSE fluctuations may exert some control on the initiation of the cycle of erosion; however, they cannot be considered as resulting in excessive erosion that negatively impacts other resources on their own, since ultimately the continuation of erosion depends on flood flows that sustain the cycle of erosion.”

This revised conclusion is based on the observed results and acknowledges that bank erosion is caused by multiple contributors at different stages of the cycle of erosion.  Importantly, it does not eliminate WSE fluctuation as one of the potential contributing factors.  It is also important to note that the report does not include a discussion of the potential loss of resistive forces such as vegetation growth at the toe of the bank due to daily WSE fluctuation, which could also contribute to the ongoing cycle of bank erosion.   Thus, the implication is that project-caused WSE fluctuations may not be the sole cause of bank erosion but that it could be exacerbating and accelerating bank erosion.

Study Conclusion #8: The study concludes in the last paragraph that “normal project operations that have changed little in several decades” (page 115 Study 2 and Study 3)

Peer Review Comment: There is no data or descriptions in the study on how operations have changed, or not changed, over time.

Recommendation #17: TransCanada should provide additional data supporting their claim that “normal project operations that have changed little in several decades”.

DELIVERABLES

This section includes our comments on the “Deliverables” as described in the RSP.  We have only commented on the deliverables from the RSP that we feel were not conducted as provided for in the RSP.

Statements from RSP: “An interim study report will be prepared after the first year of study is complete synthesizing the above deliverables into a narrative that addresses the study goals and issues raised in various study requests.  The report will be provided to stakeholders for review and comment.” (Page 25 RSP Study 2), and “The interim study report will be prepared after the first year of study is complete. The report will be provided to stakeholders for review and comment.” (Page 36 RSP Study 3)

Peer Review Comment: Interim Reports for Study 2 or Study 3 were never provided to the Erosion Working Group to review and comment.

Recommendation #18:  FERC should consider the August 1, 2016 Study 2 and Study 3 Report to be the interim report and that the Erosion Working Group’s current review of Study 2 and Study 3 Report be integrated into a revised study that the Erosion Working Group is then able to review as the final study as proposed in the RSP.

SUMMARY CONCLUSIONS

Based on our review of the Study 2 and 3 Report, our review team has made 18 recommendations as discussed earlier in this memorandum.  Most critically, we find that the Study 2 and Study 3 Report did not “ascertain the relative importance of water-level fluctuations associated with project operations in the erosion process relative to other contributing factors” as stated in the RSP and has instead stated that “trying to distinguish specific effects of normal project operations among the panoply of potential controls on bank erosion in any given location is not possible“.  A better understanding of causation should be ascertained with a different methodology such as a statistical analysis of the data collected or a bank stability model that utilizes a wider variety of geotechnical and vegetative parameters, such as geotechnical strength, maximum rooting depth, and hydraulic gradient between ground water and river water levels.

The Study Report does not consider adjacent land use as a factor (other than the presence or absence of riparian vegetation at the top of bank), and yet numerous peer-reviewed research studies have investigated and confirmed that adjacent land use has a strong role in bank stability and erosion.  Nor does the study consider the impact that daily WSE fluctuations may have on limiting vegetative growth at the toe of the river banks, although the study itself acknowledges the important role that vegetation can have on increasing bank resistance to erosion.

The Study Report does not adequately “identify the effects of shoreline erosion on other resources (e.g., riparian areas and shoreline wetlands, rare plant and animal populations, water quality, and aquatic and terrestrial wildlife habitat)” as stated in the RSP, and instead bases its conclusions on other studies that were not tasked with assessing the effects of shoreline erosion on these critical resources.

In addition, the coordination with the Erosion Working Group promised in the RSP was not conducted, and the interim reports were not delivered for review, such that the study could have been adjusted as needed to successfully complete the objectives stated in the RSP.

TransCanada should revise the Study Report or issue an Addendum to the report that includes the revisions as per the recommendations set forth in this peer review.

[1] Our review was limited to the RSP, Study 1, and the Study 2 and Study 3 Report, as well as their associated Appendixes.  No field work was conducted as part of our review, so we are not able to comment on if the observations stated in the studies accurately reflect field conditions within the project reach.  In addition, we did not review, in any detail, the numerous other studies submitted to FERC as part of TransCanada’s recent submittal.

[2] TransCanada organized stakeholders into working groups to discuss study plans and study details.  CRWC is a member of the Erosion Working Group.