April 23, 2018
Honorable Kimberly D. Bose Secretary
Federal Energy Regulatory Commission
888 First Street, NE
Washington, DC 20426
Re: Wilder Dam Project No. 1892
Bellows Falls Project No. 1855
Vernon Dam Project No. 1904
Connecticut River Conservancy Comments on Great River Hydro, LLC Study Reports filed by February 9, 2018; Request for Study Modification to Require Compliance with the RSP.
Dear Secretary Bose,
The Connecticut River Watershed Council, Inc. (CRWC), now doing business as the Connecticut River Conservancy (CRC), is a nonprofit citizen group established in 1952 to advocate for the protection, restoration, and sustainable use of the Connecticut River and its four-state watershed. We have been participating in the relicensing of the five hydropower facilities on the Connecticut River since the beginning of the process in late 2012. We have reviewed the set of Study Reports that were posted by Great River Hydro between November, 2017 and February 9, 2018. CRC attended the study report meeting held on March 8, 2018. Where necessary in our comments below, we will also refer to the Revised Final Study Report for Study 2 and 3, dated February 4, 2017.
ILP Study 2 and Study 3 Riverbank Transect and Riverbank Erosion Study Supplement to Final Study Report dated 11/15/2017
Comments based on peer review
CRC again hired consulting engineering firm Princeton Hydro (http://www.princetonhydro.com/) to conduct a peer review of the ILP Study 2 and Study 3 Riverbank Transect and Riverbank Erosion Study Supplement to Final Study Report (“Supplement”) which was submitted to the Commission by Great River Hydro on November 15, 2017. Princeton Hydro’s review is attached to this comment letter. We include some of their major conclusions below as part of our formal comments.
- The Final Study Report indicated that, “Flow velocities were measured at three impoundment erosion monitoring sites and three riverine erosion monitoring sites with an acoustic Doppler current profiler (ADCP) that measures flow velocities using the Doppler effect of sound waves scattered back from particles within the water column.” The FERC’s Determination on Requests for Study Modifications and New Studies dated July 21, 2017 (“FERC Determination”) states, “…Commission staff recommends that Great River Hydro include, in the November 15, 2017 addendum, near‐bank velocities associated with multiple water surface elevations… as measured at the six sites with ADCPs.” This information was not included in the supplemental report.
- Princeton Hydro (and CRC) request plotted cross-sections for each site with the following information shown on the same figure for each of the 21 monitoring sites: (i) annotations of erosional features (as depicted in the 2/4/17 Final Report Appendix A), (ii) water surface elevation fluctuations as measured by water level loggers, and (iii) the locations of the three sediment samples analyzed at each site in the Supplement.
- Regarding the HEC-RAS modeling, the use of a single Manning’s N, or roughness, with no differentiation between in-channel and floodplain, could produce erroneous results. The model was run in “unsteady flow” at a single flow. This is equivalent to running the model in “steady flow” and is an unusual use of the model. Our key concern is the effect of daily river fluctuations on the riverbanks, so running the model at a steady flow precludes analysis of the main source of project effects.
- Critical shear stress is not as conservative a measure as claimed in the Supplement because it does not account for cohesion, compaction, and other forces resisting entrainment.
- The presence of beaches at 18 of the 21 sites indicate that water fluctuations influence the bank similarly to the action of water in lakes and tidal areas – through repeated surface water elevation changes. Great River Hydro implies that beaches are natural. They are not natural in a riverine system. Water surface elevation fluctuations also inhibit vegetative growth on the beaches, which otherwise would contribute to the stability of banks.
- The Supplemental Study and the Revised Study do not address the role played by operational water surface fluctuations in perpetuating the bank erosion cycle. Water surface fluctuations directly contribute to bank failure resulting in sediment deposits at the toe of the bank. Without addressing the effect of water surface elevation changes at the transect sites, the Supplement does not prove that project operations are not contributing to bank erosion.
- Though the report and the final sentence of Great River Hydro’s meeting summary conclude that, “Study 2/3 results continue to show that operational flows contribute little to bank erosion,” Princeton Hydro’s peer review points out that 8 out of 21 sites showed some potential for sediment entrainment, which is a significant portion (30%) of the sites. See below for CRC’s additional comments on study conclusions.
Additional CRC comments
- The FERC Determination states that “The goals of studies 2 (Riverbank Transect Study) and 3 (Riverbank Erosion Study) were to: (1) monitor the riverbank erosion at selected sites in the project impoundments and riverine sections of the Connecticut River that are affected by the projects, (2) determine the location of erosion in areas affected by the projects and compare these locations with previously compiled erosion maps, (3) characterize the process of erosion, (4) ascertain the likely causes of erosion, [emphasis added] and (5) identify the effects of shoreline erosion on other project resources.” By avoiding any direct analysis of water surface elevation changes at the transects, Great River Hydro has not sufficiently characterized the process of erosion or ascertained the likely causes of erosion.
- The Study Plan Determination dated September 13, 2013 states, “the requested correlation [comparing water level fluctuations caused by project operations with elevations along the riverbank where there is a lack of vegetation, undercutting, or other visual signs of erosion] would provide information and would be useful to identify the causes of erosion (§5.9(b)(5) and (6)). Besides water level fluctuations, other causes of erosion include land use practices, ground water seeps, gullies, and high flows. A stated objective of the study is to ascertain the likely causes of erosion [emphasis added] at various locations. Project operations would be a likely cause of erosion where visible signs of erosion closely track project-caused water level fluctuations…” [emphasis added]. Additionally, the Study Plan Determination states, “As a result, we recommend modifying study 3 to correlate visible indicators of erosion with project-caused water level fluctuations [emphasis added] at the 20 transect locations…” Project caused water fluctuations include daily surface water elevation changes at the dam. The Revised Final Study Report and Supplement have failed to adhere to the Study Plan Determination.
- The FERC Determination requires that, “Great River Hydro file an addendum… that includes an analysis of estimated critical shear stress, near-bank velocity, and the potential correlation of these factors with project operation at the 21 monitoring sites. This discussion should include a table for each monitoring site that lists critical shear stresses and near-bank velocities with respect to water surface elevations corresponding to project operation…” [emphasis added]. Project operations include daily fluctuations in surface water elevation at the dam, not just changes in flows with the dam held at a single elevation. Great River Hydro chose to take sediment samples based on modeled surface water elevations at transects while maintaining no surface water fluctuation at the dam. CRC contends that this was not what was asked of them in the FERC Determination and this limited interpretation of the FERC Determination does not support the goals of the study.Not only did the analysis for the supplemental report not involve river fluctuations, but the dam elevations used to run the velocity and sheer stress analysis do not correspond with dam operation elevations typically used for those flows. Figures in the Pre-Application Documents (PADs) dated October 2012, for example Figure 2.5-1 in the Wilder PAD, provide “normal generation ranges” for each impoundment, and it also shows the reservoir profile operation for elevation at each dam. The table below summarizes the dam elevations used in the supplemental report for the “minimum,” “average operational,” and “capacity” flows vs. the flows those elevations correspond to under normal operations according to the PAD. The dam elevation used for most of the Vernon Dam analysis is particularly odd, since it lies outside of the normal operational range. According to the PAD, each dam is held at higher elevations for flows within the facility’s operational control, and for higher flows, each dam’s elevation is lowered. That is the opposite of what was done for the analysis in this Supplement. Therefore, the dam elevations used for the analysis do not appear to reflect typical operation elevations for those flows, potentially calling the entire analysis into question.
||Flows used in supplemental report (cfs)
||Dam elevation used in Supplement, Appendix A
(NAVD88 ft msl)
|Elevation converted to NGVD29 ft msl
||Flow (cfs) corresponding to the NGVD29 elevation in PAD Figure 2.5-1
384.6 (W09 and W12 max only)
291.2 (B09 max only)
219.6 (V06 max only)
|Unknown, outside of normal operation range
- Princeton Hydro’s peer review of the Revised Final Study Report dated 5/15/2017 noted, “The data presented in Table 5.8‐1 [of the Revised Final Study Report] actually show that velocities increase between 36% and 400% during these periodic operational drawdowns, resulting in velocities significantly in excess of the threshold velocity for sediment entrainment later discussed in Section 6.1. The data presented in Table 5.8‐1 therefore suggests that periodic operation drawdowns, in preparation for high flows, could regularly mobilize sediment at the toe of the streambank at 9 of the 13 monitored impoundment cross sections.” We had hoped that because of FERC’s request for additional analysis, the Supplement would shed some more light on this observation, but Great River Hydro instead set up their model runs for the supplemental analysis to completely avoid this issue altogether. They held the impoundment at the same elevation, and for the sites closest to the dam, the model used a higher impoundment level to run the “max” elevations, which is directly contrary to their practice, according to the PAD, of lowering the impoundment elevations for higher flows.
- The FERC determination stated that, “Great River Hydro include… an analysis of the stratigraphy at the 21 monitoring sites, including, at a minimum, a discussion of any potential correlation between erosive features (e.g., notches, undercutting) and soils present within normal operation ranges” [emphasis added]. Normal project operational ranges would include daily fluctuations in surface water elevation (SWE) at the dam and the resulting fluctuations at transect sites at various points along the river. The license allows surface water elevations at the dam to fluctuate by several feet. By maintaining the SWE at the dam at the same elevation they are not actually modeling the operations of the dam. Both variables, SWE fluctuations and velocity of water, need to be considered.
- CRC is concerned that many of the transect sediment samples were taken at elevations that do not correspond to where the surface water elevations would actually fall on the bank. Slide 28 presented during the Updated Study Report meeting clearly gives the impression that the sediment station at the upper part of the bank corresponds to the “maximum flow,” the mid part of the bank corresponds to the “medium” flow, and the lower part of the bank corresponds to the “minimum” flow. This does not seem to be how it was actually done, though. For instance, the Supplement states, “Similarly, at some sites, especially impoundment sites just upstream of a dam (e.g., W12), the WSE for the 3 operational conditions were essentially at the same elevation since the nearby dam WSE remained unchanged for all operational flows considered.” Additionally, the sediment sample elevations for many of the sites either fall completely outside of the median WSE fluctuation or only one sampling site falls within that area of the bank. As far as we can tell, the soil samples have no particular connection with the river flows and dam elevations used in the model, and moreover, some don’t include samples within typical operational ranges. See attached graphs for B03, V03, V06, W03, and W10 depicting where we think the soil samples were collected, given the information provided in the Supplement [note: we could only use the sample elevation to determine sample station location because the “sample station (ft)” corresponded to a horizontal distance from the hydraulic model which differ from the horizontal distances shown in figures Appendix A of the Final Revised Study]. We have also plotted the logger data for W10 as an example of where the sediment samples fall in relation to daily fluctuations – we note for this figure that the Supplement Appendix A lists the “max” elevation of 383.4 as “dry” for the 700 and 5,000 cfs model runs, therefore giving no velocity readings, but according to the logger graph included here, listing the max elevation as dry at 5,000 cfs does not appear to be accurate.
- Also of concern is the fact that we have no way to know actual or average surface water elevation fluctuations for December to May of most years since actual SWE for those months was not provided due to the difficulty of logger placement in winter. As mentioned above, the validation of the model using surface water elevations at the 6 ADCP sites was not included in the supplement. We request that this information be provided, and it include maximum historic operational surface water elevation changes at the dam and resulting surface water elevation changes at the transect sites for various flows.
- The analysis of entrainment of average sediment particle size is problematic. It may very well be that the average sediment particle size is high because clays and fines have been removed from the bank due to surface water fluctuations. This would skew the velocity needed to move sediments to be a higher threshold velocity. Additionally, focusing on the velocity needed to move the average size particle ignores the erosion of up to 50% of the sediment material, including the loss of clays and fines and resulting reduced bank cohesion. Ignoring the impact on clays and fines also ignores the possibility that the structure of the bank is being destroyed.Shear stress (and entrainment) is based on the description of moving materials away from the base – it is not what causes the material to be at the base. CRC contends that shear stress at various operational flows is not the issue. At issue is the change in cohesion due to repeated wetting and drying of the banks as a result of water surface elevation changes. The velocity of water draining out of the bank as water surface elevations go down and sediments are removed was not considered. By not considering cohesion or the process of upper bank erosion, the Supplemental Study primarily examines the mechanism of moving sediments that have already eroded from the bank.
- The FERC determination states that, “Great River Hydro include near bank velocities associated with multiple water surface elevations… as measured at the six sites with ADCPs. For the remaining 15 sites… the average velocity associated with multiple water surface elevations as calculated by the HEC-RAS model. If, possible, Great River Hydro should include a discussion or estimate of the near-bank velocity or these 15 sites based on available data.”During the study plan meeting on March 8, 2018, Lissa Robinson (GEI Consultants) stated that, “Sub-critical flow – in the riverine flow you would have downstream flow. Sub-critical flow is in the pools where flow might go upstream; for each 10 feet by 10 feet cell in the HEC-RAS model you would have velocity that could flow in multiple directions. It could pick up and model an eddy if it did exist.” While GRH provided tables for each transect site, it is not clear if the velocity listed is “near-bank” or average velocity. Additionally, based on Lissa Robinson’s comment it is not clear the direction of the flow of velocity. Is it downstream, based on an eddy, or upstream?
- The Supplemental Report states, “Colluvial material derived from erosion higher on the bank still covered the stratigraphy at the base of the banks at many of the monitoring sites as was the case during the two years of monitoring from 2013 to 2015.” The question is not why the colluvial material hasn’t moved (and erroneously, thus erosion is not taking place). It is instead, “why is there colluvial material at the toe of the bank?” If the study had answered that question it might have “ascertained the likely causes of erosion” as required as a goal of the study.
- The Revised Final Study Report also states, “The degree of change at the [ADCP] monitoring sites does not appear to be related to flow velocities as some of the sites with the highest flow velocities experienced no or little change during the two-year monitoring period… Similarly, some of the sites with the lowest flow velocities experienced the greatest amount of change during the two years of monitoring as at the Bellavance Site. The comparison between flow velocity and documented change at the monitoring sites shows no strong relationship and indicates that other factors [emphasis added]… may also exert some control on the location of bank changes.” Those other factors may well be the loss of fines and clays from repeated water surface elevation fluctuations. The Supplement and Revised Final Study Report did not address this.
- The licenses for Wilder, Bellows Falls and Vernon were issued in 1979, prior to the completion of the USACE Connecticut River Streambank Erosion Study: Massachusetts, New Hampshire and Vermont (1979). The 1979 license states, “The New Hampshire Fish and Game Department recommended that NEPCO [New England Power Corporation, the previous owners of these dams] be required to stabilize bank conditions within the impoundment area. The Department contends that fluctuation of the reservoir level has caused serious bank erosion and resultant siltation in the Connecticut River. Intervenors including For Land’s Sake, have also raised this issue. Over 100 protests to the issuance of a long-term license to NEPCO, prior to the completion of the US Army Corps of Engineers Study have been received on the subject of erosion… In our order we denied For Land Sake’s motion that we not issue a license for the Wilder Project until the erosion study was complete.”The USACE Connecticut River Streambank Erosion Study: Massachusetts, New Hampshire and Vermont, 1979 study (“1979 Army Corps Study”) states, “Evaluation of forces causing bank erosion verifies the relative importance of causative factors. In descending order of importance they are: shear stress (velocity), pool fluctuations, boat waves, gravitational forces, seepage forces, natural stage variations, wind waves, ice, flood variations, and freeze-thaw. Analysis of the causes of bank erosion shows that these causes can be subdivided into those that cause general bank erosion and those that cause upper bank erosion. Tractive forces exerted by flowing water cause general bank erosion, with their maximum attack occurring at about two-thirds of the depth below the water surface. Hence, even if the upper bank is stable or stabilized, the flow can erode the lower bank causing failure of the lower and upper banks. Forces such as wind waves, boat waves, pool fluctuations, ice, etc., are the most common causes of upper bank erosion… In time, a berm or beach is formed… Furthermore, limited control of upper bank erosion can be achieved by limiting pool fluctuations associated with hydropower development…” [emphases added]. CRC contends that the focus on instream velocity and entrainment only addresses part of what is going on. The Final and Supplemental reports for Studies 2 and 3 have still not addressed pool level fluctuations and the resulting effects of upper bank erosion. Focusing on the entrainment and movement of already eroded and non-cohesive sediment is not proof that project operations do not contribute to the overall erosion cycle.
- The 1979 Army Corps study states, “The magnitude of the energy gradient has been altered by the low head hydropower dams… the analysis of the stability of the system must consider the changes imposed on the slope of energy gradient by the systems of dams. The system no longer operates as a free-flowing alluvial channel. Its energy gradient and the velocity have been reduced except for those reaches above the influence of the pools.” Additionally, the Revised Final Study states, “NRCS’ (2007) publication on thresholds for small channel design recommends a maximum permissible velocity of 1.5 feet per second (ft/s) for fine sand in clear water without any detritus but 2.5 ft/s in water carrying colloidal silts as higher velocities are needed to transport silt and clay, because of their cohesiveness, than fine sand.” Hence, basing the velocity threshold on the NRCS thresholds for small channel design may not be appropriate.
- The 1979 Army Corps study says on page 67, “Comprehensive literature surveys reveal that numerous experienced engineers and geologists have concluded that 90-99% of all significant bank erosion occurs during major flood events. These observations are not based upon concept or theory, but on field observation.” [emphasis added]. We went into this relicensing process knowing that major flood events cause changes in river morphology, and we did not need several years of study to confirm this. As we said in our comments from July 15, 2013 on the Preliminary Study Plan for Study 2, “The problem of erosion is not just a matter of high flows and ice out scour. There is legitimate concern that daily reservoir level fluctuation causes piping of water in and out of a saturated bank, piping that would be an important contributor to the erosion problems landowners are experiencing in the impoundment areas.” Great River Hydro, and FirstLight as well, have both focused on the erosion processes related to high flow events, ignoring the impact that daily river fluctuations from project operations contribute to bank erosion (including instability that can then lead to bank failure during high flow events).
- The Kleinschmidt Lower Connecticut River Shoreline Survey Report (2011) states that, “Sand and silt particles that make up the bank and bed material along the river erode most readily. Also, decreases in shear strength of the soil bank material may lead to failure. This is especially true where swelling of fine soil materials from absorption of water increases groundwater pressure within the bank, and soil creep (downhill slope movement) weakens the bank…. Bank slumping, sometimes described as mass failure or collapse can occur from various mechanisms, but is most commonly a result of rapid draw down of stream flow following a prolonged period of bank-full flow (high water or flood flows with a relatively rapid reduction in flow) resulting in saturation of bank material.” Even though Great River Hydro paid for the Kleinschmidt study, with the conclusions of their Supplemental Study they still are ignoring the impact that daily river fluctuations from project operations have on bank erosion.
- We have included our notes from the March 8, 2018 study report meeting to be added to the record to supplement the summary provide by Great River Hydro in order to provide additional detail in regard to specific questions asked and the flow of discussion.
CRC recommendations and conclusions
Based on the peer review and our own analysis, CRC continues to believe that Studies 2-3: (1) were conducted in violation of the Revised Study Report (RSP) dated August 14, 2013 and approved with modifications from FERC on September 13, 2013; (2) did not follow several recommendations from FERC’s Determination on Requests for Study Modifications and New Studies dated July 21, 2017; and/or (3) otherwise reached conclusions that the science, data or evidence do not support.
In accordance with 18 CFR §5.15(a), CRC recommends GRH do the following:
- As mentioned above, the validation of the model using surface water elevations at the 6 ADCP sites was not included in the supplement. We request that this information be provided, and that it include maximum historic operational surface water elevation changes at the dam and resulting surface water elevation changes at the transect sites for various flows.
- Prepare figures showing cross-sections for each site with (i) annotations of erosional features (as depicted in the 2/4/17 Final Report), (ii) water surface elevation fluctuations as measured by water level loggers, (iii) the water surface elevations corresponding to the three discharges analyzed in the Supplement, and (iv) soil sample locations used in the supplement.
- Provide graphs that show velocity across the span of the river at transect sites as shown in Slide 27 in the Study Report meeting presentation.
- Great River Hydro indicated that they have a gradation of sediment size for all samples taken. Please provide a table showing the percentages of particle sizes in the corresponding sediment samples and what particle size could be moved by various near bank velocities.
- Otherwise address all concerns described in these comments and the Princeton Hydro peer reviews.
The issue of erosion continues to be widespread in the project area and worsens year by year. These issues were brought to the attention of FERC by a significant number of river citizens almost 40 years ago during the last relicensing process and were not addressed at that time or since. We request that the FERC recognize its public trust responsibility and ensure that erosion control and streambank stabilization figure prominently in the relicensing of these facilities.
Great River Hydro’s conclusion that project operations do not cause erosion has not been proven and is not supported by the evidence provided in numerous studies. The Supplemental study was not designed in a way that reflects normal operational conditions and ultimately only examined the velocity needed to entrain an average sediment particle. CRC contends that the studies conducted by GRH have not adequately considered or identified the possible causes of erosion. At this point in the process, we believe the licensees of the Connecticut River projects are not going to adequately look at operational effects on bank erosion. Consequently, CRC requests that FERC conduct a robust review of the Great River Hydro and FirstLight erosion studies, including the raw data from all underlying models used (HEC-RAS, River2D, BSTEM). Impoundment fluctuations are widely understood to contribute to erosion. Both companies will have to provide ways to avoid, minimize, and mitigate the impact. CRC recommends a publicly warned site visit by qualified FERC personnel to examine the eroding riverbanks first-hand. The FERC site visits that took place during the fall of 2012 as part of the relicensing scoping process are now more than five years in the past, the tours did not look at erosion sites close up from the land nor cover much of the impoundment, and many FERC staff currently involved in the relicensing were not present at the tours. CRC is glad to help coordinate this site visit if needed.
ILP Study 18: American Eel Upstream Passage Assessment
A primary goal of the study was to determine how well temporary eel ramps might work when the fish ladder is not functional. During the study period, the ladder was open three weeks longer than usual, until August 7, and this may have confounded the study results.
We appreciate the ongoing support and enhancements that Great River Hydro is making to provide eel passage. As upgrades are made to the ladder, pit tag studies should be conducted to evaluate the efficacy of changes made. Additionally, the ladder should be open to allow for the full seasonal upstream and downstream migration.
ILP Study 21: American Shad Telemetry Study – Vernon
A goal of Study 21 was to evaluate downstream passage routes and survival. It would be helpful to have analysis that shows routes specific to project operation states, and associated survival. For example, what are common routes and survival rates when there is spill vs. when there is not spill? Similarly, what are routes and survival rates when there are certain turbines operating vs. not operating? Without this information there is not enough data to inform operational scenarios that support the success of downstream migration.
In addition to the comments provided above for both Study 18 and Study 21, please note that CRC also supports the comments submitted by the natural resource agencies, including but not limited to, the New Hampshire Fish and Game Department (NHFGD), the Vermont Fish and Wildlife Department (VTFWD), and the U.S. Fish and Wildlife Service (USFWS).
We appreciate the opportunity to provide comments on the studies submitted by February 9, 2018. I can be reached at email@example.com or (802) 258-0413.
- Princeton Hydro peer review dated May 15, 2017
- River Stage profiles for B03, V03, V06, W03, W10
- Logger Data for W10
- Urffer notes from March 8, 2018 Study Report Meeting