June 8, 2016


Honorable Kimberly D. Bose Secretary

Federal Energy Regulatory Commission

888 First Street, NE

Washington, DC 20426


Re:       Northfield Mountain Pumped Storage Project No. 2485-063

Turners Falls Project No. 1889-081

Connecticut River Watershed Council response to FirstLight responses dated May 31, 2016


Dear Secretary Bose,


On May 31, 2016, FirstLight submitted its responses to comments received on the 13 study reports and 2 addendums filed on March 1, 2016.  In this letter, the Connecticut River Watershed Council, Inc. (CRC) is writing to FERC to indicate that several of our issues were not addressed in FirstLight’s responses, and remain as significant concerns.  Don Pugh assisted us in preparing this letter.

Stakeholder involvement on studies to be re-done

We have learned in FirstLight’s response dated May 31, 2015 that FirstLight plans to re-do portions of the Odonate Study (3.3.10) and the Ichthyoplankton Study (3.3.20) based on conference calls that have occurred between FirstLight and agency staff.  CRC requested studies, has attended meetings, reviewed reports, and commented on these studies.  We clearly have demonstrated an interest in these two studies.  In the spirit of the Integrated Licensing Process (ILP), we believe that all stakeholders should be given an opportunity to participate in, or at least be notified, when there are meetings, emails, and drafts of plans that lead to re-doing a particular study.

Use of historical data and the hydraulic study

CRC commented on FirstLight’s plans for using flow and operational data from 2002 and 2009, respectively, in Study 3.2.2 in our letters dated May 2 and May 3, 2016.  FirstLight did not respond in their May 31, 2016 letter.  Stakeholders have commented on the use of specific years of river fluctuation data for certain studies, and the need to look at historical data to be able to understand conditions that will represent more than a single year of study.  We believe that is what the Hydraulic Study, 3.2.2, was geared toward.  The Revised Study Plan (RSP) for the Study 3.2.2 listed seven studies that were to evaluate how water level fluctuations associated with project operations impact various resources.  Those studies are listed in the background section of Study 3.2.2 RSP as Study 3.1.2 (erosion causation study), 3.3.6 (shad spawning), 3.3.10 (odonate study), 3.3.13 (littoral zone fish habitat and spawning), 3.3.17 (tributary and backwater habitat), 3.5.1 (wetland habitat and special status species), and 3.6.6 (recreation and land use).  At the time of the RSP for 3.2.2, FirstLight had not committed to doing Study 3.3.20, and we would include that study in this list as well.  CRC requests that FERC provide guidance about using the historical data set and the Hydraulic Study to look at operation effects rather than individual years that vary from study to study.

Shad Spawning Study 3.3.6


In their cover letter dated May 31, 2016, FirstLight responded to CRC’s May 25, 2016 request that Study 3.3.6 be repeated by saying the request was not filed in a timely manner and the issues that caused the request are addressed in responses USFWS-1 and USFWS-2.  The issues that concern CRC are not addressed in either of FirstLight’s responses (USFWS-1 or -2), or in any other FirstLight response.  USFWS-1 is related to the surveys in the canal and bypass reach.  CRC did not ask, in its study request, that those reaches be studied again.  CRC requested that the study only be repeated in the reach below Cabot.  USFWS-2 addresses habitat duration curves and the area available for spawning under different discharges.  While this issue is also a concern, this is not the reason CRC requested the study be repeated.  CRC asked that the study be repeated because the flow changes as evidenced by the USGS Montague gage do not appear to reflect what was intended in the FERC-approved study plan (18 C.F.R. § 5.15(d)(1)).


Flow Discrepancy


FirstLight’s response to comment CRC-1 (from our letters dated May 2 and corrected May 3, 2016) referred to the response to USFWS-7 and a corrected Table 4.3-1A included in Attachment B to Study 3.3.6.  The response to USFWS-7 does not respond to the concern raised in CRC-1 (or a similar concern raised by the USFWS in USFWS-7 or the NMFS in NMFS-5) that the gage at Montague does not show the changes in operations noted for each of the tests (Site ID 8 to 21).  To better illustrate our comment and concern, Table 1 below is adapted from FL’s corrected Table 4.3-1A to highlight the flow issues noted by the CRC, USFWS, and NMFS.


Table 1 below lists discharge at the Montague gage for the start of the before-and-after counts, the difference between the two discharges, the number of units operating before and after the change, and the change in the number of units.  The ‘Difference from Expected’ column is how different the expected change based on the unit change was from the USGS change.  Differences in the opposite direction are noted in red.  The sites highlighted in green are in close proximity to the Montague gage and should reflect, in real time, the change in generation.  Flows in the Deerfield remained stable during the period of spawning surveys.  The orange highlighted columns are ours added to columns taken directly from Table 4.3-1A in Attachment B.


The April 29, 2015 RSP for Study 3.3.6 stated that, “Based on historic generation data at Cabot Station, most changes will be +/- 10 Mw (2,288 cfs) followed by 20 Mw (4,576 cfs) changes.”  The USGS flows indicate that no flow changes greater than 1,800 cfs occurred during the study.  The way the units have been reported does not match up with the resulting flows, according to the Montague gage.  Though the RSP determined that a 10-20 MW change happens 29% of the time, the study apparently never experienced that flow change, as demonstrated in the USGS gage flows.  Therefore, we continue to believe that the portion of the study below Cabot station should be repeated to observe flow changes that are representative of operations and what was intended in the RSP.



Table 1: Modified from Table 4.3-1A
Date Instantaneous River Flow Montague USGS Number of Units    Site ID
Before After Gage Change Before After Change Expected flow change based on units change (cfs) Difference From Expected (cfs)
5/26/2015 8,310 8,150   -160 1 2 Up +2,280 2,440 8
5/26/2015 8,830 9,000   170 2 1 Down -2,280 2,450 9
5/27/2015 11,000 9,240   -1,760 0 1 Up +2,280 4,040 10
5/28/2015 8,690 8,190   -500 1 0 Down -2,280 1,780 11
5/28/2015 7,710 9,240   1,530 2 3 Up +2,280 750 12
5/28/2015 9,760 9,150   -610 3 2 Down -2,280 1,670 13
6/9/2015* 12,500 12,700   200 4 4   Same 200 14
6/9-10/2015 16,000 16,200   200 6 4 Down -4,560 4,760 15
6/10/2015 21,300 20,900   -400 6 4 Down -4,560 4,160 16
6/10-11/2015 19,400 18,400   -1,000 4 6 Up +4,560 5,560 17
6/16/2015 20,400 20,400   0 6 4 Down -4,560 4,560 18
6/17/2015 17,600 18,700   1,100 4 6 Up +4,560 3,460 19
6/17/2015 15,800 15,600   -200 4 2 Down -4,560 4,360 20
6/17-18/2015 14,100 13,500   -600 2 4 Up +4,560 5,160 21
6/22/2015* 26,400 n/a     6 n/a   n/a   22
* Ambient conditions did not permit operators to change Cabot generation.
Red numbers indicate flow changes opposite change in unit operations.
Green highlights are sites in close proximity to the Montague gage.



Splash Count Analysis


FL did not address CRC’s analysis of the before-and-after splash counts using the Wilcoxon signed-rank test that shows a statistically significant difference between the before and after splash counts (letter dated May 2 and corrected May 3, 2016).  We assume that this is because FL did not disagree with the analysis or the conclusion.


Rather, to explain the difference in splash counts, FirstLight relied on the Durbin-Watson test of autocorrelation and the trend in shad splash counts over time (photoperiod – FL letter May 31, Attachment F) to assert that time of day and time of year, respectively, are the controlling reasons that splash counts declined in Phase 2 study samples.


The Durbin-Watson test is incorrectly applied, as not only did time pass between the before and after splash counts, but there was also a change in generation that also affected shad behavior.  It is more likely that changing flow decreased counts than did the passing of time.  As for the increase in photoperiod, the number of shad available to spawn was not constant through the six week period of study.  More shad passed Holyoke in May than June during the 2015 migration season (Figure 1 below), consequently more spawning would occur earlier in the study period than at the end.  By June 18th, the last day of Phase 2 counts, very few fresh shad were in the study area.


Figure 1: Shad passage at the Holyoke Dam, 2015



Field Methods


In CRC’s March 25, 2016 information request, we asked for information on boat launch locations and times.  FirstLight’s response dated April 22, 2016 indicated this information would be coming on May 31, 2016.  FirstLight’s response to NMFS-1 for study 3.3.6 says, “The RSP did not require recording launch sites, boat tracks, or the amount of time spent in each location. “

We are surprised that FirstLight’s field methods did not include recording launch sites or boat tracks.  The meeting minutes from a March 24, 2015 stakeholder meeting at the Hampton Inn in Greenfield indicated discussion related to making sure the same track wasn’t repeated at the same time on each survey date

[1].  The final RSP for this study dated April 29, 2015 stated, “Sampling will be conducted randomly to ensure the results are not biased by visiting the same site at the same time of day every time.”  Without any information provided in the report or responses, CRC is unable to determine if the study was conducted in accordance with the final RSP (see 18 CFR. § 5.15(d)(1)).



Study 3.3.10 Assess Operational Impacts on Emergence of State-Listed Odonates in the Connecticut River

For CRC, many questions remain as to the details of the 2016 field work and supplemental report.

  • Will the supplemental report summarize historic information from the Turners Falls impoundment, as required in the original RSP?
  • Where is the supplemental field work going to take place?
  • In response to MADFW-10, FirstLight says, “FirstLight will use the hydraulic model to evaluate potential effects of water level changes on odonates at additional locations in the TFI downstream from the Montague USGS Gage.” This makes no sense, as the TFI is located upstream of the USGS gage.
  • In response to CRC-1, it is unclear what odonate information will be used in this analysis – only the data collected for this study, historical data, or a combination? It is entirely unclear to us that FirstLight is actually going to provide an analysis of operational effects in the Turners Falls impoundment, and what data will be used for this analysis.  We can see no information in the responses related to this.
  • In response to CRC-2, we question why data from earlier studies could not be used.

CRC requests that a revised study plan for the 2016 work be posted publicly and provided to stakeholders for further comment.


Study 3.3.20, Ichthyoplankton Study


CRC commented in our letter dated May 2, 2016 and corrected on May 3 that FirstLight did not include river flow in its analysis of entrainment.  FirstLight’s cover letter of May 31, 2016, states:


CRC’s comments on Study No. 3.3.20 reflect a misunderstanding of the study. The study is based on the density of organisms and flow into the generation facility; Vernon discharge is not a component of the entrainment estimate, and river flow is never a component in this type of entrainment estimate.


In their study determination letter of January 22, 2015 the FERC recommended that river discharge be included in the analysis of entrainment:


Including river discharge in the analyses of ichthyoplankton density and entrainment rates would require minimal additional cost (section 5.9(b)(7)) and could inform the development of license requirements (section 5.9(b)(5)).  Therefore, we recommend that FirstLight include river discharge in its analyses of ichthyoplankton density and entrainment rates.


Because FirstLight did not include river discharge, the report is incomplete and should be revised.  CRC’s request for discharge from the Vernon project is reasonable and should be a part of the study report (18 C.F.R. § 5.15(b)).


FirstLight is repeating the entrainment study in 2016.  According to an email dated May 19, 2016 from Christine Tomichek of Kleinschmidt in response to an inquiry from Don Pugh, the study began on May 10th and a study plan was to be provided on May 30.  No study plan for 2016 entrainment study was provided in FL’s May 31, 2016 letter.   Response to USFWS-1 states that, “The [2016] study will be conducted similar to the 2015 study except sampling will occur once per week at whatever pumping flow occurs that evening and no attempt will be made to manipulate the flows as required in last year’s study  FL also agreed to conduct offshore sampling (three tows) every night an entrainment collection is made.”  Stakeholders have had no opportunity to comment on the 2016 plan.  In their report for the 2016 study, FirstLight should include river flow as a component of their analysis.


FL did not respond to CRC’s request for a more complete description of the calculation of adult equivalents, including an analysis of variability.  The entire explanation of adult equivalence is one short paragraph of 104 words and two tables.  A more complete explanation, including variance, should be included in a revised study report.


Study 3.6.1  Recreation Use/User Contact Survey


One of the goals of Study 3.6.1 was, “Determine the amount of recreation use and demand at the Turners Falls and Northfield Mountain recreation sites.”  Similarly, a goal of TransCanada’s Study 30 was, “Obtain information about existing recreation use and opportunities, access, and present and future use estimates for sites within and in riverine sections between the projects.”  In developing the RSP’s for all the relicensing studies, FERC stressed that the studies did not need to be done with the exact same study methods, because there are multiple methods at getting to our questions.  However, if the goal is to determine recreation use at all facilities, then stakeholders would of course want results that can be compared across studies.  If in this case, numbers cannot be compared, then something is wrong with the whole process.


FirstLight has explained that it used a combination of traffic counter data and calibration count data to arrive at user numbers per site.  This makes sense.  However, the rationale for using something like 10.9 traffic counts per visit in the summer vs. 13.7 in the fall is not provided.  Vehicles likely circle within a boat launch site as boats are brought in and out of the water.  But the tube counters were placed at the entrance and exits of these sites, and so it is not clear why on average a vehicle would pass in and out of a site multiple times, and more so in the fall.  CRC has the calibration data spreadsheets, and there were no data collected that would indicate the same vehicle crossing over the tube counter multiple times (more than in and out) a visit.  Additional details on study methods and rationale would be helpful.


Andrea Donlon

River Steward

[1] “Andrea asked as part of the Phase 1 work what does random site selection mean. Chris noted that the randomness is in time and that they will not monitor the same spot at the same time. Don asked how many sites can be assessed per night. Bryan believed it would be around 4-5 sites at night, but cautioned that is subject to various conditions. Melissa noted that based on historic data below Cabot Station, there are 4-5 sites where they have found shad spawning multiple times.”