CRC comments on Updated Study Report for Northfield Mountain and Turners Falls Dam

November 13, 2015

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

Comments on the Updated Study Report (USR) dated September 14, 2015

Comments on the 2015 Initial Study Meeting Summary dated October 14, 2015

Dear Secretary Bose,

The Connecticut River Watershed Council, Inc. (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. Below are our comments on the September 14, 2015 Updated Study Report (USR) and October 14, 2015 Initial Study Meeting (ISM) Summary submitted by First Light.

Please note that our comments on recreation studies were sent under separate cover in a joint letter with the Appalachian Mountain Club (AMC).

3.1.1 2013 Full River Reconnaissance Study

CRC continues to believe that Study 3.1.1, the Full River Reconnaissance (FRR), was completed in a way that differed from the Approved Study Plan. In particular, the study did not use the study plan’s and the QAPP’s definition of “stable” in making study conclusions. “Stable” is defined in the FRR as “riverbank segment does not exhibit types or indicators of erosion.” Using their FirstLight’s data and assessment of each segment, we will stress that only 43.3% of the banks meets their definition of stable.

CRC and the Connecticut River Sreambank Erosion Committee (CRSEC) have not yet responded to FirstLight’s August 4, 2015 “Letter to FERC on response to information regarding sites for bank stabilization and preventative maintenance for P-1889 and P-2485.” We regard this as a license compliance issue and will submit comments at a later date.

3.1.2 Northfield Mountain/Turners Falls Operations Impact on Sediment Transport

As a member of CRSEC, CRC was involved in drafting the letter commenting on study 3.1.2 submitted by CRSEC. We include the letter as CRC Attachment 1 in order to endorse the comments and include them as part of our own.

The list of references in the December 15, 2014 response to stakeholder comments and in the USR are missing some of the references cited in the Field 2007 study and additional ones cited in the letter from the University of Illinois attached to the CRSEC’s letter on the Initial Study Report (ISR) dated November 14, 2014. These references should be considered for review and inclusion in the study report.

3.2.2 Hydraulic Study of Turners Falls Impoundment, Bypass Reach and below Cabot

One of the goals of the study per the revised study plan (RSP) was to provide water surface elevation (WSEL) data and mean channel velocity information at the model transects to inform other studies. No information on mean channel velocity was included in the report – no discussion of the data, findings under a range of flow and project operating conditions, or implications for use in other models and other studies. This information should have been included in the 3.2.2 Study Report.

The RSP for Study 3.2.2 stated under Task 3 that, “The HEC-RAS model will be operated for a given flow and the WSELs measured at FirstLight monitoring locations will be compared to the model results.” Why does the model not estimate WSELs at the tailrace logger or the Turners Falls Dam in Figures 5.1.2-1 through 11? As we could see from the logger data, the upstream of Northfield logger did not behave quite the same as the tailrace logger in 2013 and 2014. CRC is interested in the degree of fluctuation in the vicinity of the Northfield tailrace for a number of reasons. The current Northfield Mountain license sets parameters for the fluctuation range of the “lower reservoir” – the Connecticut River – based on river elevation as measured at the Turners Falls Dam. It is critical that stakeholders see model outputs for these two locations.

Figure 2.2.1-2 in the report shows a flow duration curve at the Montague USGS gage for 2012, 2013, 2014, and the period of record. Flows in 2014 were roughly similar to the period of record; 2013 was vastly different (lower flows).

Why aren’t there flow duration curves for the water logger data, “naturally routed flow” data and Vernon gage data?

August 24-September 3, 2014 was the calibration time period shown in Figures 4.1.3-2 through 11, but Appendix B figures for the loggers does not seem to depict the tailrace logger data for that time period. Why?

We are concerned with both the magnitude and frequency of changes in the WSEL. The study report presents the model results with a brief discussion of magnitude of the WSEL changes, but does not discuss the frequency of changes in WSEL. It would be informative to see the unsteady-state model runs depicted at a 15-minute time step interval rather than the 12-hour intervals shown in Figures 5.1.2-1 through 5.1.2-11 on pages 5-10 through 5-20. The time interval used flattens or masks the effects of the project operations on WSEL. Data is available for a 15-minute time step and the model results should be shown at this time interval as well as other shorter intervals. The effect of project operations on bank stability, for example, may be better determined at a finer scale than the graphs shown.

There was no discussion of the model results relative to the 2013 and 2014 water level logger and flow data for Vernon and Northfield Mountain (Appendices A and B) and the impacts on WSELs. The following comments reflect our observations and questions after looking through the graphs in the appendices.

• If Turners Falls logger levels are at all influenced by what is being moved to the canal, or spilled at the dam, then that information should have been included in the graphs in Appendices A and B.
• There are discrepancies between observed water levels at the Turners Falls dam and the Northfield Tailrace during high flow periods. For example, April 8-10, 2014 shows a flow event in which the tailrace loggers and upstream loggers all respond to high flows coming downstream from Vernon, whereas the Turners Falls logger shows a decrease. Again, from April 11-15, 2014, there is a gradual increase in flows from Vernon and the tailrace and upstream loggers all show water levels rising, whereas the Turners Falls dam logger shows steady or decreasing levels.
• May 19-20, 2014: When Vernon is spilling water and flows from Vernon are steady to declining, while Northfield generatrs, river levels go up from the tailrace to the state line. Levels at the French King gorge and Turners Falls Dam stay steady, but then increase when Northfield is withdrawing. What is the explanation for the pattern at TF dam?
• It appears from the logger data for May 1-6, 2014, that if Vernon operations are steady, when Northfield pumps (negative), water levels at the tailrace dip but are steady at the dam. When Northfield generates (positive), levels drop at the dam but are higher at the tailrace.
• August 6-11, 2013: Northfield tailrace logger shows river levels dip down low when Northfield is pumping. No other logger responds quite in this way, which is to be expected. Interestingly, the logger does not respond in an opposite way when Northfield is generating. There is a positive response at the tailrace logger to generation between August 29-30, 2013, however.
• The Northfield tailrace logger shows some erratic behavior September 1-3, 2013, and is oddly at a level consistently higher than many loggers downstream and upstream of it. The logger does not respond to Northfield pumping like the August dates did. Is this accurate data?
• July 6-11, 2014: Vernon is relatively steady, and it shows the fluctuations of the entire pool in response to Northfield pumping and generating, even at the logger located upstream of Stebbins. On the other hand, the Northfield tailrace logger in this graph does not respond to Northfield pumping as it did in 2013. Why?
• The summertime logger at the Northfield tailrace responds in a certain way but there are some questions about the 2013 data. In 2014, the tailrace logger was out of service for much of the summer. CRC is concerned that there is a lack of good data for the tailrace area that were input into the model.
• It would be useful to see the water level logger data shown at a finer resolution so that the pulses or waves of water, especially during the generation cycle of Northfield Mountain could be evaluated. Similarly, during the pumping cycle, the change in WSEL is likely to be less smooth than what is shown in the figures and pulses in WSEL as the water is drained from the impoundment should be shown. The scale of the figures (24-hour period) masks these changes, particularly in 2014 when the pages had a portrait elevation. CRC has the Excel files for the logger data but did not have the opportunity to do a more thorough analysis before the comment deadline.

Observations:

• It appears that sites in the mid-Turners Falls pool experience summer-time sub-daily 2-4-ft fluctuations (the model shows 2.5-foot fluctuations in Figure 5.1.2,-Scenario 4). This can impact riverbank erosion and recreational use of the river. Flows generally decrease at midnight until mid morning or mid day, then steadily increase during the latter half of the day. This can have impacts on streambank erosion and recreational users downstream.
• Based on the August 11-16, 2012 graph downstream of the Turners Falls dam (Appendix C), peaking flows out of the Cabot units can result in 5-ft sub-daily fluctuations in Montague and 4-foot subdaily fluctuations at the Sunderland Bridge in the middle of the summer. Flows rapidly decrease at midnight until mid morning or mid day, then steadily increase during the latter half of the day. This can have impacts on recreational users downstream.
• Peaking hydropower generation in the Deerfield River does not have a dramatic effect on flows in the Connecticut River downstream of the Turners Falls Dam.

CRC request. CRC requests that FirstLight provides channel velocity information. We recommend an addendum to the report that includes more discussion and analysis of observed logger data. We recommend that the model runs be re-run showing the Northfield tailrace and the Turners Falls dam location, and that the graphs be scaled to show the same number of days, but with more space per time segment.

3.3.2 Evaluate Upstream and Downstream Passage of Adult American Shad

The RSP for this study indicated that “An initial report summarizing existing data evaluated in Task 1 will be prepared.” Though it is not noted in the USR Meeting Summary, I asked about this deliverable at the meeting and was told they are “still working on it.”

CRC request: The study report for study 3.3.2 is proposed to come out by September 1, 2016. CRC would like to know if this means that the Task 1 report and the report for Tasks 2 and 3 are being merged into a single report, or if two separate reports are still proposed and if so, what is the schedule for completion. CRC it may be helpful to have the Task 1 report complete before the planned workshop proposed for this winter (at the USR meeting, we were told this may happen in December or January) to review and discuss results.

3.3.3 Evaluate Downstream Passage of Juvenile American Shad

During the USR meetings, we learned that one of the units for Northfield Mountain is out for service until February. This should be noted as a study variance. We aren’t sure why this wasn’t mentioned in any of the USR reports.

3.3.5 Evaluate Downstream Passage of American eel

During the USR meetings, we learned that one of the units for Northfield Mountain is out for service until February. This should be noted as a study variance. We aren’t sure why this wasn’t mentioned in any of the USR reports.

3.3.7 Fish Entrainment and Turbine Passage Mortality Study

During the USR meetings, we learned that one of the units for Northfield Mountain is out for service until February. This should be noted as a study variance. We aren’t sure why this wasn’t mentioned in any of the USR reports.

3.3.17 Assess the Impacts of Turners Falls Project and Northfield Mountain Project Operations on Tributary and Backwater Area Access and Habitat

This is a completed study.

CRC could not see how the study objective of measuring changes in available habitat and water quality in lower tributaries resulting from Project-related water level fluctuations was achieved in this report. In no way could we discern how the habitat in each tributary varies with the sub-daily fluctuations from Project-operations, during each season, and what that would mean for aquatic organisms.

The RSP considered low water conditions to be 176.0 feet msl or as close as is practical at the time of the study in the Turners Falls impoundment. The study report says that the data collected on August 12, 2014 are representative of an “abnormally low” impoundment level (178.3 ft at the TF Dam), coupled with a low Vernon discharge of 1,536 cfs, created by conditions that were manipulated by an upstream TransCanada study. We aren’t sure why the study plan set a low flow level that was impossible to achieve. The locations measured on August 5 and August 11, 2014 had impoundment levels that were 2-5 ft higher, so results are not quite consistent with one another.

In looking at Table 5.1-1 of the report, some of the water surface elevations (WSELs) did not vary much from season to season, however. For example, at Mill Brook, the spring measurement was taken at WSEL 182.67, the summer at 181.06, and the fall at 182.03 ft. At Pine Meadow Brook, located close to the Northfield Mountain tailrace and presumably susceptible to project operation fluctuations from Northfield, the spring measurement was taken at WSEL 180.53, the summer at 179.15, and the fall at 182.13. We had expected field surveys to represent a range between high water conditions to low water conditions. And how the tributary conditions vary under a subdaily fluctuation of 4 feet or so, we don’t know.

The summer field days also do not represent “low flow” tributary conditions. August and early September of 2014 was characterized by flows in nearby tributaries as being higher than the median flows for those dates. See below for USGS gage data for Priest Brook (in the Millers basin) in Winchendon, MA and the Ashuelot River in Hinsdale, NH.

CRC request: FirstLight create an analysis that shows habitat impacts from Project operations at each of the study sites during the three seasons that were studied. New data may need to be collected to gather seasonally-representative flow conditions and site-specific changes under Project operations.

3.3.18 Impacts of the Turners Falls Canal Drawdown on Fish Migration and Aquatic Organisms

This is a completed study.

As the report notes in section 1.2, two other studies (3.3.3 and 3.3.5) will address the first objective to the study, which was to assess whether juvenile American shad and American eel abundance in the canal increases leading up to the time of its closure for the drawdown. These studies are not due out until September 1, 2016 and March 1, 2017, respectively.

Our comments below focus on the other goals of the study.

Comments on Goal 2: Evaluating the level of mortality for juvenile sea lamprey

The study report shows in Tables 4.2-1 and 4.2-2 that in 32 quadrats, only 6 live ammocetes and 1 live transformer were found on day 1 and 5 live ammocetes and 0 transformers were found on day 2. No dead organisms were observed on either day. CRC believes that the low numbers of lamprey indicates something wrong with the methodology.

The Revised RSP submitted with the ISR from Sept. 2014 states that, “The final number and placement of quadrats will be decided prior to the drawdown in consultation with the resource agencies and Conte Lab researchers. These data will be used to extrapolate counts for the entire area of suitable habitat within each zone.” The ISR contained copies of consultation with agencies, and comments from USFWS and MA DFG (Natural Heritage staff) were included and addressed. We do not see that consultation included staff from the Conte Lab. CRC is concerned that the quadrats that were chosen either did not capture sea lamprey habitat, or the timing was wrong.

In trying to assess how the final quadrat locations were chosen, CRC notes that the revised RSP submitted as Appendix to ISR for 3.3.18 said that in Task 1, “A field crew of experienced biologists will systematically traverse each of the zones in a meander survey fashion recording observations of estimated number of each species encountered.” We could not find a record of a meander survey having taken place, or any information about any observation, in the final report. Based on the report, we cannot determine if the quadrat locations are representative of ammocete or transformer habitat in the canal. We corresponded with fisheries biologist Boyd Kynard about this, and have provided his memorandum and curriculum vitae (CV) as CRC Attachment 2. In his comment 1, he estimates tens of thousands of sea lamprey larvae and thousands of juvenile in the canal that were “high and dry” that likely died as a result of the drawdown. Dr. Kynard’s comment 2 addresses a few of the problems with the quadrat approach.

In assessing if the timing was wrong, we see that in the methods section, the report states that results labeled as “Day 1” in the report represent field work conducted during the hours of 1300-1800 on Monday and 0800-1430 on Tuesday. We do not know what time drawdown commenced, or at what time some of the areas started to be dewatered. However, according to Dr. Kynard’s comment 3, Dr. Kynard observations are that most sea lamprey leave their burrows and crawl to the surface within 2 hours after water has receded, and “all” emerge within 8.5 hours after an area has been dewatered. There is no way to tell in the report what time the quadrats were sampled, or when the sites were picked, but according to Dr. Kynard’s observations, the surveys could be done visually without digging, as long as they are done in a specific time window. He was out between 8-10 AM on Monday, 3-5 hours earlier than the FirstLight team started.

Please also see comments from Dr. Kynard on the mudpuppy surveys in CRC Attachment 2.