DRAFT REPORT

Email comments to mjwiley@umich.edu

 

 

  [download a MS WORD version of this document]

 

A RETROSPECTIVE ANALYSIS OF PRE- AND POST-RELICENSING

FLOW REGIMES ON THE LOWER MUSKEGON RIVER

 

by

 

B. Rudy, M. Wiley, and P. Richards

 


 

 


 The Croton hydroelectric facility on the Muskegon River, Newaygo County

 

 

{This report is based on a thesis submitted

by Bruce C. Rudy

in partial fulfillment of the requirements

for the degree of

Master of Science

(Natural Resources and Environment)

University of Michigan

August 2003}

 

 

 

 

 

ABSTRACT

 

            In 1994, three hydroelectric facilities on the Muskegon River (Rogers, Hardy, and Croton Dams) were relicensed by the Federal Energy Regulatory Commission.  Relicensing terms reflected, in part concerns regarding the impacts dam operations were having on the downstream river ecosystem.  These concerns included fisheries habitat degradation from increased water temperature, decreased dissolved oxygen content, and elevated rates of bank erosion associated with “peaking” operations.  The 1994 FERC agreement mandated changes in operations of the dams with the goal of significantly changing hydrologic regime of the Muskegon River below.

            The focus of this study was to compare the flow regimes associated with the pre- and post-relicensed hydropower facilities and discuss how the relicensing of these facilities has in fact altered the downstream hydrologic regime and hydraulic potential.  The three criteria used in the analysis of the altered flow regimes were event peak delay and attenuation impacts, water balance and flow frequency impacts, and alterations to the river’s hydraulic energy and sediment transport potential.

            Results of the study suggest that relicensing the hydroelectric facilities has in fact altered the water balance and flow regime of the lower river albeit in ways not necessarily anticipated in the FERC relicensing plan.  Furthermore, hydraulic energies in the Muskegon River appear to have been significantly reduced.  On the other hand, storage delay was minimally altered by the change in operations and hydrograph attenuation was almost nonexistent.

 


ACKNOWLEDGEMENTS

            We would like to thank everyone who helped to assure the success of this study.  Funding came from the Great Lakes Fisheries Trust as a part of the GFLT Muskegon River Initiative. This study was based upon data collected, and generously provided by the USGS, NOAA, the Michigan DNR, and Research staff from the School of Natural Resources and Environment, University of Michigan. 

 

 


TABLE OF CONTENTS

ABSTRACT. ii

ACKNOWLEDGEMENTS. iii

TABLE OF CONTENTS. iv

LIST OF FIGURES AND TABLES. iv

INTRODUCTION.. 1

METHODS. 4

The Study Sites and Their Settings. 4

Rogers Facility. 4

Hardy Facility. 5

Croton Facility. 5

Regulatory History. 5

Data Collection. 6

Pre-relicensing Streamflow Data. 6

Post-relicensing Streamflow Data. 7

Precipitation and Evapotranspiration Data. 7

Data Analysis. 7

Water Balance. 7

Flow Duration Curves. 8

Double Mass Analysis. 9

Power Estimations. 9

RESULTS. 10

Hydrograph Delay and Attenuation. 10

Water Balance and Flow Regime. 10

Flow Duration Analysis and Total Water Balance. 10

January and February Flow Duration and Water Balance. 11

March and April Flow Duration and Water Balance. 12

May through August Flow Duration and Water Balance. 13

September through December Flow Duration and Water Balance. 14

Double Mass Analysis. 15

Energy and Erosion Potential 16

DISCUSSION.. 17

Hydrologic Regime and Energy Impacts. 17

Ecological Impacts. 19

LITERATURE CITED.. 38

LIST OF FIGURES AND TABLES

Table 1.  Physical characteristics of the Rogers, Hardy, and Croton impoundments. 20

Table 2.  Operating characteristics of the Rogers, Hardy, and Croton impoundments. 20

Table 3.  Land use characteristics of the Rogers, Hardy, and Croton impoundments. 21

Table 4.  USGS gauging station information for selected sites on the Muskegon and Little Muskegon Rivers. 21

Table 5.  Flow duration percentiles and percent difference for pre-relicensing (1990-1993) input and output and post-relicensing (1996-2000) input and output for the Muskegon River. 21

Table 6.  Hydrograph flows for input, output, ET, and reservoir flux (δ) for 1991 and 2000 for the Muskegon River impoundments. 21

Table 7.  Flow duration percentiles and percent difference for pre-relicensing (1990-1993) and post-relicensing (1996-2000) input and output for the Muskegon River in January and February. 22

Table 8.  Average hydrograph flows for input, output, ET, and reservoir flux (δ) normalized to output for 1991 (pre-relicensing) and 2000 (post-relicensing) for the Muskegon River impoundments. 22

Table 9.  Flow duration percentiles and percent difference for pre-relicensing (1990-1993) and post-relicensing (1996-2000) input and output for the Muskegon River in March and April. 22

Table 10.  Flow duration percentiles and percent difference for pre-relicensing (1990-1993) and post-relicensing (1996-2000) input and output for the Muskegon River in May through August. 22

Table 11.  Flow duration percentiles and percent difference for pre-relicensing (1990-1993) and post-relicensing (1996-2000) input and output for the Muskegon River in September through December. 22

Table 12.  Pre-relicensing (1991) and post-relicensing (2000) stream power and percent difference for the Muskegon River from the tailwaters of Croton Dam to Muskegon Lake. 23

Figure 1.  The Muskegon River watershed surrounding the Rogers, Hardy, and Croton dams. 23

Figure 2.  Linear representation of the inflows and outflows of the Rogers, Hardy, and Croton impoundments on the Muskegon River. 25

Figure 3.  The Rogers hydroelectric facility on the Muskegon River, Mecosta County. 25

Figure 4.  The Hardy hydroelectric facility on the Muskegon River, Newaygo County. 25

Figure 5.  The Croton hydroelectric facility on the Muskegon River, Newaygo County. 25

Figure 6.  Scatter plot and linear regression for correlation of measured stream flow at Evart gauging station to Big Rapids gauging station. 25

Figure 7.  Scatter plot and linear regression for correlation of measured stream flow at Little Muskegon at Morley gauging station to Little Muskegon at Oak Grove gauging station. 26

Figure 8.  Scatter plot and linear regression for correlation of measured stream flow at Newaygo gauging station to Croton gauging station. 26

Figure 9.  Water balance components for Muskegon River hydrologic equation. 27

Figure 10.  Fifteen-minute hydrograph for the year 1991.  Input flows determined using the Evart to Big Rapids correlation  and output flows determined using the Newaygo to Croton correlation. 27

Figure 11.  Fifteen-minute hydrograph for 2000.  Input flows measured at the Big Rapids gauging station and output flows measured at the Croton gauging station. 28

Figure 12.  Precipitation exceedance probabilities for pre-relicensing (1990-1993) and post-relicensing (1996-2000) years. 28

Figure 13.  Flow duration curve for pre-relicensing (1990-1993) and post-relicensing (1996-2000) input and output for the Muskegon River. 29

Figure 14.  Flow duration curve for pre-relicensing (1990-1993) and post-relicensing (1996-2000) input and output in January and February for the Muskegon River. 29

Figure 15.  Hydrograph for the Muskegon River input, output, ET, and reservoir flux (δ), including dS/dt, small tributary inputs, and GW, for January and February 1991. 30

Figure 16.  Hydrograph for the Muskegon River input, output, ET, and reservoir flux (δ), including dS/dt, small tributary inputs, and GW, for January and February 2000. 30

Figure 17.  Hydrograph for the Muskegon River input, output, ET, and reservoir flux (δ), including dS/dt, small tributary inputs, and GW, normalized to output for January and February 1991. 30

Figure 18.  Hydrograph for the Muskegon River input, output, ET, and reservoir flux (δ), including dS/dt, small tributary inputs, and GW, normalized to output for January and February 2000. 31

Figure 19.  Flow duration curve for pre-relicensing (1990-1993) and post-relicensing (1996-2000) input and output in March and April for the Muskegon River. 31

Figure 20.  Hydrograph for the Muskegon River input, output, ET, and reservoir flux (δ), including dS/dt, small tributary inputs, and GW, for March and April 1991. 31

Figure 21.  Hydrograph for the Muskegon River input, output, ET, and reservoir flux (δ), including dS/dt, small tributary inputs, and GW, for March and April 2000. 32

Figure 22.  Hydrograph for the Muskegon River input, output, ET, and reservoir flux (δ), including dS/dt, small tributary inputs, and GW, normalized to output for March and April 1991. 32

Figure 23.  Hydrograph for the Muskegon River input, output, ET, and reservoir flux (δ), including dS/dt, small tributary inputs, and GW, normalized to output for March and April 2000. 32

Figure 24.  Flow duration curve for pre-relicensing (1990-1993) and post-relicensing (1996-2000) input and output in May through August for the Muskegon River. 33

Figure 25.  Hydrograph for the Muskegon River input, output, ET, and reservoir flux (δ), including dS/dt, small tributary inputs, and GW, for May through August 1991. 33

Figure 26.  Hydrograph for the Muskegon River input, output, ET, and reservoir flux (δ), including dS/dt, small tributary inputs, and GW, for May through August 2000. 33

Figure 27.  Hydrograph for the Muskegon River input, output, ET, and reservoir flux (δ), including dS/dt, small tributary inputs, and GW, normalized to output for May through August 1991. 34

Figure 28.  Hydrograph for the Muskegon River input, output, ET, and reservoir flux (δ), including dS/dt, small tributary inputs, and GW, normalized to output for May through August 2000. 34

Figure 29.  Flow duration curve for pre-relicensing (1990-1993) and post-relicensing (1996-2000) input and output in September through December for the Muskegon River. 34

Figure 30.  Hydrograph for the Muskegon River input, output, ET, and reservoir flux (δ), including dS/dt, small tributary inputs, and GW, for September through December 1991. 35

Figure 31.  Hydrograph for the Muskegon River input, output, ET, and reservoir flux (δ), including dS/dt, small tributary inputs, and GW, for September through December 2000. 35

Figure 32.  Hydrograph for the Muskegon River input, output, ET, and reservoir flux (δ), including dS/dt, small tributary inputs, and GW, normalized to output for September through December 1991. 35

Figure 33.  Hydrograph for the Muskegon River input, output, ET, and reservoir flux (δ), including dS/dt, small tributary inputs, and GW, normalized to output for September through December 2000. 36

Figure 34.  Double mass plot for cumulative precipitation versus cumulative output for pre-relicensed (1990-1993) and post-relicensed (1996-2000) data. 36

Figure 35.  Double mass plot for cumulative input versus cumulative output for 1970 – 1973, 1974 – 1977, 1978 – 1981, 1982 – 1985, 1986 – 1989, pre-relicensing (1990 – 1993) and post-relicensed data (1996-2000). 36

Figure 36.  Scatter plot of input versus output for pre-relicensing (1990-1993) and post-relicensing (1996-2000) operations. 37

 


INTRODUCTION

            The Muskegon River, which runs through central and western Michigan, is 220-miles long, fed by 94 tributaries, and has a watershed of approximately 2,634 square miles.  The river begins in a tiny stream in Roscommon County called Big Creek, which drains to Higgins then Houghton Lakes via a channel known locally as “the cut” (Alexander, 1994c).  Once the river exits Houghton Lake and Reedsburg Dam, it slowly grows and makes it way through relatively flat upper reaches.  Just above the town of Evart, the Muskegon’s gradient becomes steep and the river’s velocity increases.  It is in this middle, higher gradient portion of the river where three large hydroelectric facilities (Rogers, Hardy and Croton) intercept the Muskegon’s path to Lake Michigan.  Below this section of the Muskegon, the river flattens out once again as it meanders through a freshwater estuary and ends its journey in Muskegon Lake.

            While the Muskegon is one of the best known recreational rivers in the State of Michigan, many are concerned that it functions more like a series of shorter rivers, rather than the one continuous 220-mile river it is (Seelbach, 1999).  The history of the Muskegon River reveals much about how it became a “fragmented” river.  The first sawmill appeared on the river in 1837 in the town of Newaygo and the first dam was built in Big Rapids in 1866.  From 1837 to 1906, about 30 billion board feet of wood was floated down the river.  Between 1900 and 1908, three more dams were built in the middle portion of the Muskegon River: Newaygo Dam (1900), Rogers Dam (1906), and Croton Dam (1907).  The last sawmill closed in 1910 and in 1931, Hardy Dam, the largest earthen dam at the time, was built.  The last dam to be built on the Muskegon River was the Reedsburg Dam in 1940 (Alexander, 1999a).  In 1966, the Big Rapids Dam was dismantled and in 1969, the Newaygo Dam was removed. 


The Rogers, Hardy, and Croton dams, and their effects both positive and negative, have however remained, and continue to be a source of controversy.

            The role of dams in river ecosystems has gained increasing attention from the scientific community in recent decades (Chin et al., 2002).  While dams have provided numerous benefits such as flood control, water supply, recreational activities, and hydroelectric power, they have also altered flow regimes and ecologically fragmented rivers systems, as well as often causing geomorphic and ecological  change (Graf, 1992; Dynesius and Nilsson, 1994).  The Muskegon River mainstem dams have offered increased recreational opportunities on the river and have offered flood control and electricity generation to the residents of Mecosta and Newaygo Counties.  But, Rogers, Hardy, and Croton Dams have also had real impacts on the Muskegon River ecosystem, including increased fish mortality due to entrainment in the power plants, blockage of fish passage, and alterations of channel habitat.

Annual fish entrainment in 1997 at the three dams was estimated to be 301,583 fish with a total annual mortality was 44,042 fish.  The river fragmentation that resulted from these structures has had both positive and negative effects.  The dams have acted as a barrier to pest species, such as the Sea Lamprey, keeping them from invading the upper reaches of the river.  But, the dams have also acted as a barrier to the movement of desireable migratory fish.  Changes to river water quality have in the past included elevated temperature and decreased dissolved oxygen content downstream of the dams, and in the view of some increased bank erosion due to regular cycles of saturation and drainage associated with peaking operations.  Peaking hydroelectric operation is defined as operations where a dam holds water and only releases it at peak times when electricity prices are the greatest (Geis, 1982).  Impacts attributed to this operational strategy include losses in downstream fish habitat due to high water velocities during peak discharge (Q); increased sediment erosion and lowering habitat diversity; disruption of normal downstream movement of carbon materials important to aquatic life; and creation of lake environments within the river system which resulted in lower fish productivity and shifts in fish communities favoring lake species (O’Neal, 1997).

Many of these concerns were raised and addressed during the relicensing of the Rogers, Hardy and Croton hydroelectric facilities in 1994.  The Federal Energy Regulatory Commission (FERC), working with the Michigan Department of Natural Resources (MDNR) and Consumers Power Company, the owners of the dams, undertook a study on the environmental impacts of the dams on the river prior to the relicensing.  The study determined that altering the operations of the facilities from peaking to run-of-river might mitigate many of the negative impacts of the dams.   Run-of-river hydroelectric operation is defined as operations where the flow out of a reservoir is equal to the flow into a reservoir (Geis, 1982).

In this report, we compare the flow regimes associated with the pre- and post-relicensed hydropower facilities and discuss how the relicensing of these facilities has altered the downstream hydrologic regime and hydraulic potential.  To accomplish this, we examined the effects the dams have had on pre- and post relicensing hydrographs in regards to peak delay and attenuation.  Additionally, we compared the pre- and post-relicensing impacts to the river’s water balance and flow regime.  Lastly, we examined the effects of the altered operations on the river’s hydraulic energy and erosion potential.


METHODS

The Study Site

            This study focuses on the three impoundments within the Muskegon River watershed in Mecosta and Newaygo Counties, Michigan (Figures 1 and 2).  The Muskegon River, from Reedsburg Dam in Roscommon County, Michigan, travels approximately 134 miles and drops about 280 feet before reaching the Rogers hydroelectric facility (Figure 3).  Once the Muskegon River passes through Rogers Dam, it enters Hardy Pond.  Hardy Pond is 25 miles long and terminates at the Hardy hydroelectric facility (Figure 4).  Water flowing over Hardy Dam enters Croton Pond.  Croton Pond is about 6 miles long and terminates at the Croton hydroelectric facility (Figure 5).

            Through the length of the impoundments, there are several small tributaries that enter the mainstem through the reservoirs.  The total catchment area of these streams is approximately 131 square miles.  The catchment area of the Muskegon River measured at the tailwaters of the Croton dam is approximately 2,313 square miles.  The tributaries represent about 5.7% of the entire Muskegon River catchment at this point.

Rogers Facility

            The Rogers facility is the uppermost impoundment of three impoundments within this study.  It is located at river mile (RM) 83.7 on the Muskegon River.  A full description of the impoundment’s physical characteristics has been provided by Consumers Power (Table 1).  Built in 1905-1906 by the Grand Rapids Muskegon Power Company (GRMPC), the hydroelectric facility was historically and continues to be a run-of-river facility (Consumers Power Company, 1994b).  Detailed information on the post-


relicensing operating characteristics of Rogers Dam and the land use surrounding the Rogers impoundment were provided by Consumers Power (Tables 2 and 3).  The impoundment is maintained at 861.3 feet above sea level and is allowed to fluctuate plus/minus one foot (Consumers Power, 1994b).

Hardy Facility

            The Hardy impoundment is between the Rogers and Croton facilities.  It is located at river mile (RM) 58.7 on the Muskegon River.  The largest of the impoundments, Hardy Reservoir’s physical characteristics were provided by Consumers Power (Table 1).  Built in 1929-1931 by Consumers Power (CP), the Hardy facility was and continues to be operated as a peaking facility (Consumers Power Company, 1994c).  Detailed information on the post-relicensing operating characteristics of Hardy Dam and the land use surrounding the Hardy impoundment were provided by Consumers Power (Tables 2 and 3).

Hardy impoundment is lowered from nine to twelve feet in January and February of every year to prepare to store river flows anticipated in March and April from the annual snowmelt and thaw.  This spring draw down allows the Hardy impoundment to absorb the snowmelt and runoff from the surrounding watershed upstream of the facility (Sluka, 2002).  During May through December, the impoundment is maintained at 822 feet above sea level and is allowed to fluctuate plus/minus one foot (Consumers Power, 1994c).

Croton Facility

            Croton pond is the last of the impoundments within this study.  The Croton hydroelectric facility is located at river mile (RM) 52.8 on the Muskegon River.  A full description of the impoundment’s physical characteristics has been provided by Consumers Power (Table 1).  Built in 1906-1907 by GRMPC, the Croton facility was historically a run-of-river facility.  The relicensing of the Croton hydroelectric facility resulted in a change in the operation of the dam to “re-regulate” or “smooth out” the peaking operation at Hardy Dam (Consumers Power Company, 1994d).  As a result of this change in operations, the three hydroelectric facilities on the Muskegon River now operate in a run-of-river manner.  Detailed information on the post-relicensing operating characteristics of Croton Dam and the land use surrounding the Croton impoundment were provided by Consumers Power (Tables 2 and 3).  The impoundment is maintained at 722 feet above sea level and is allowed to fluctuate plus/minus one foot (Consumers Power, 1994d).

Regulatory History

            The FERC issued the original operating licenses for the Rogers (FERC #: 2451-004), Hardy (FERC #: 2452-007), and Croton (FERC #: 2468-003) hydroelectric facilities on May 24, 1968 (Consumers Power Company, 1994a).  The licenses allowed the Muskegon River Dams to operate in a manner that promoted electricity production at the time of its highest price (i.e. peaking operations).  The Michigan Department of Natural Resources claimed that operating the facilities in this manner has been shown to put undue stress on the ecosystem downstream from the facilities.  In general, streambed scouring, channel armoring, bank and island erosion, and stream channel degradation often result from peaking operations (Grant et al., 2003).  On the Muskegon, bank erosion was and still is common.  It was attributed in pre-relicensing debates to peaking operations.

            New operating licenses were issued by the FERC for the Rogers, Hardy, and Croton hydroelectric facilities on July 15, 1994 (Consumers Power Company, 1994a).  The new licenses required the Muskegon River hydroelectric facilities, as a total unit, to operate in a way that resembled the natural flow regime of the river (i.e. run-of-river).  Specifically, the Croton facility, which had historically operated as a run-of-river facility, would be forced to alter its operations to re-regulate the flow from the Hardy hydroelectric facility and allow the river to re-equilibrate to its natural flow regime (Consumers Power Company, 1994a).

            The goal of the new licenses was to operate the Croton hydroelectric facility so that the flows leaving the Croton facility were approximately equal to the inflows of the upstream Rogers’ facility plus the inflows from the Little Muskegon River (Figure 2).  The Rogers facility is operated in run-of-river mode, therefore, flows at the peaking Hardy facility are re-regulated by the Croton impoundment to return the Muskegon River to its natural flows.  To accomplish this, the sum of the projected mean daily discharge from the Hardy Reservoir plus the inflow from the Little Muskegon River are adjusted for time of passage and water accretion and are then measured at the downstream Croton gauge.  This comparison is made on an hourly basis.  Time of passage and water accretion from Hardy Dam to Croton Dam was estimated to be 0.5 hours and 2.6%, respectively.  Time of passage and water accretion from the Little Muskegon to Croton Dam was estimated to be 3 hours and 3.2%, respectively (Consumers Power, 1998).  During the draw down and refill of the Hardy facility in the spring, the Croton facility releases the projected mean daily discharge from the Hardy impoundment plus the inflow from the Little Muskegon River (Consumers Power, 1994a).

Data Collection

            Discharge data used for this thesis came from the United States Geological Survey (USGS) gauging stations above and below the three impoundments (Table 4).  Due to the changes in gauging stations as a result of the FERC relicensing, two sets of gauges were used, one set for the pre-relicensing data and one set for the post-relicensing data.  Linear regression was used to predict what the flows at the pre-relicensing gauging stations might have amounted to at the post-relicensing gauging stations during the pre-relicensing flow regime (Figures 6, 7, and 8).  The slope of the regression equation was used to correlate the flows.

Pre-relicensing Streamflow Data

            Stream discharge for 1970 – 1993, was collected at the Evart gauging station (USGS #: 4121500), Little Muskegon at Morley gauging station (USGS #: 4121900), and Newaygo gauging station (USGS #: 4122000).  Detailed information about these gauging stations was provided by the USGS (Table 4).  Average daily flows in cubic feet per second (cfs) were used for the years 1970 – 1993.  The term pre-relicensing years is defined as the time period 1990 – 1993.  Fifteen-minute flows in cfs were used for the year 1991 in the peak delay and attenuation analysis.

Post-relicensing Streamflow Data

            Stream discharge for 1996 – 2000, was collected at the Big Rapids gauging station (USGS #: 4121650), Little Muskegon at Oak Grove gauging station (USGS #: 4121944), and Croton gauging station (USGS #: 4121970).  Detailed information about these gauging stations was provided by the USGS (Table 4).  Average daily flows in cfs were used for the years 1996 – 2000.  The term post-relicensing years is defined as the time period 1996 – 2000.  Fifteen-minute flows in cfs were used for the year 2000 in the peak delay and attenuation analysis.

Precipitation and Evapotranspiration Data

            Precipitation data for the area surrounding the three impoundments was collected at the Big Rapids Waterworks station (NOAA COOP ID: 200779).  Daily precipitation in hundredths of inches was used for this site for both the pre- and post-relicensed years.

            Calculation of evapotranspiration (ET) was determined using Penman’s equation.  An Excel-based version of Penman’s equation written by R.L. Snyder (2000) from the University of California was used for this calculation.  The calculation required monthly maximum and minimum temperatures, average wind speed, average dew point temperature, and net solar radiation.  This data, with the exception of monthly net solar radiation, was collected at the Muskegon County Airport station (NOAA COOP ID: 205712) for both the pre- and post-relicensed years.  Due to the lack of consistent monthly net solar radiation data, the maximum monthly net solar radiation was assumed for the pre- and post relicensed years.

Data Analysis

            The data were analyzed using a variety of standard hydrologic techniques.  Components of the water balance were attenuation for both the pre-relicensing and post-relicensing years.  Flow duration curves, hydrographs, and double mass plots were developed for pre- and post-relicensed time frames.

Water Balance

Data available to estimate water balance for the reservoir system (Rogers to Croton) included gauged input along the mainstem of the Muskegon above Rogers Dam and from the Little Muskegon River (InputUSGS), gauged output along the mainstem below Croton Dam (OutputUSGS), and ET estimates (see above).  Input from ungauged smaller tributaries draining into the reservoirs (Inputtribs), and change in reservoir storage (dS/dt), and groundwater (GW) (Figure 9).  A water balance equation for the combined reservoir system can be written as:

dS/dt = InputUSGS + Inputtribs +Gwin – (OutputUSGS +Gwout + ET)                      (1)

The InputUSGS component used for water balance estimates was the sum of the Evart and Little Muskegon at Morley gauging station for the pre-relicensing time period and was the sum of the Big Rapids and Little Muskegon at Oak Grove gauging stations for the post-relicensing time period.  The OutputUSGS component of the water balance was the Newaygo gauging station for the pre-relicensing time period and the Croton gauging station for the post-relicensing time period.  Linear regression was used to correlate the flow data from the pre-relicensing gauging stations to the post-relicensing gauging stations (Figures 6, 7, and 8).

Several of the water balance components could not be measured directly.  The sum of dS/dt, GW, and Inputtribs were unmeasured and termed reservoir flux, designated gamma (d).

d = ±dS/dt ± GW + Inputtribs                                        (2)

Since from the water balance (equation 1),

OutputUSGS = InputUSGS ± d - ET                                  (3)

Then,

d = OutputUSGS – InputUSGS + ET                                  (4)

Based on the run-of-river operating characteristics of the impoundments, dS/dt was assumed to be near zero for the months May through December.  The 12-foot draw down of the Hardy impoundment during January and February equals a volume of 1.3 billion cubic feet, and dS/dt was assumed to be constant over the draw down period (e.g. volume of draw down/number of days of draw down) (Consumers Power Company, 1994a).  The requisite fill-up of the Hardy impoundment during March and April was calculated in a similar manner.  The value associated with dS/dt during the draw down and refill was determined to be approximately 14.7% of the output measured at the Croton gauge.  Based on the contributing drainage area, the input associated with the ungauged tributaries (Inputtribs) should be approximately 5.7% of the output measured at the Croton gauge.  GW is presumably the remaining flux associated with d and represents the majority of the “reservoir flux” within the impoundments.

Because a time of passage was estimated by Consumers Power and in this study (see Results), the input component of the water balance was shifted two days forward to account for the travel time of the river from the gauging stations above the impoundments to the gauging station below the impoundments (Consumers Power Company, 1998).

Flow Duration Curves

            Flow duration curves used in the analysis of pre- and post-relicensed operations of the facilities offered a perspective on the change in the input and output flows that occurred through the three facilities from 1990-1993 to 1996-2000.  Average daily flow data was sorted in descending order, ranked from one to n, and then given an exceedance probability by using the formula:

                                                            T = (n + 1)/m                                                   (5)

Where n is the flows of record and m is the rank magnitude of that flow (Allan, 1995).

Double Mass Analysis

            Ambiguities due to both delayed and/or non-linear hydrologic response can be avoided by use of a double mass analysis (Linsley et al., 1982).  Double mass plots were used in the analysis of pre- and post-relicensed operations of the facilities to offer a perspective on how the output of the pre- and post-relicensed systems responded to precipitation and to input.  Discharge (output) and input were measured in cfs and converted to cfd.  Precipitation was measured in hundredths of inches and converted to inches.

Power Estimations

            The result of the hydrologic regime change at the three hydroelectric facilities on the Muskegon River was an altered hydrologic energy.  The amount of energy available for geomorphic work is proportional to both the amount (mass) of water being moved in the channel (discharge (Q)), and the rate of fall (slope (S)) of the channel (Knighton, 1998).  The equation that represents the power of a river is:

                                                            Ω = 10QS                                                        (6)

Calculations of this equation were applied to both pre- and post-relicensing operations of the Muskegon facilities to determine the change in power of the system.  Discharge (Q) was determined from the fifteen-minute data in cfs and converted to cubic meters per second (cms).  Slope (S) was determined from USGS gauge #: 4121970 at the tailwaters of Croton Dam to the mouth of Muskegon Lake (USGS gauge #: 4060102), which has fluctuated an average of 4.92 ft. since 1990 (Boutin, 2000).  The distance between these two gauges was determined to be 47 river miles (Alexander, 1999b).  Based on the fluctuation, the slope was calculated to be 0.00042 to 0.00044; 0.00043 was used for the calculation.  The results of the power equation were in kilowatts per foot (kW ft-1) and were converted to MWh ft-1 for the pre- and post-relicensing operations in 1991 and 2000, respectively.


RESULTS

Hydrograph Delay and Attenuation

            Consumers Power Company (1998) estimated the time required for the passage of water from Big Rapids to Croton Dam to be approximately 48 hours.  Inspection of graphs of input, from Big Rapids, and output, from Croton Dam, support the Consumers Power estimate (Figures 10 and 11).  In the pre-relicensing operations the peak delays were approximately 2 days (Figure 10) and in the post-relicensing operations, the peak delays were again approximately 2 days (Figure 11).

            There was attenuation of storm peaks between the input measured at Big Rapids and the output measured at Croton Dam.  On the contrary, the input versus output for both pre- and post relicensing years suggests that there was often significant accretion occurring between the input measured at Big Rapids and the output measured at Croton Dam (Figures 10 and 11).  This accretion implies that additional volumes of water entered the Muskegon River system between Big Rapids and Croton Dam.

Water Balance and Flow Regime

Flow Duration Analysis and Total Water Balance

The pre-relicensing years witnessed a greater amount of precipitation than did the post-relicensing years (Figure 12).  The pre-relicensing flows exceeded the post-relicensing flows for all exceedance probabilities, likely due in part to greater precipitation in the pre-relicensing years (Table 5, Figure 13).  The pre-relicensing output exceeded the pre-relicensing input for the majority of flows.  The difference in median flow (Q50) for the pre-relicensing input and output was 27.56%.  Post-relicensing input and output flows were more closely aligned and very close at high flows (Q5)


The difference in Q50 post-relicensing input and output flows was 4.25%.  The differences in shape of the pre- and post-relicensing input and output suggest that the post-relicensing flows were managed carefully at all flows.  Alternatively, pre-relicensing input and output resembled a less consistent flow regime that managed the difference in output to input at approximately 10% or more.

The total water balance for the pre-relicensing year 1991 and post-relicensing year 2000 offered a more detailed perspective on the components that comprise the Muskegon River within the impoundments (Table 6).  Again, the difference in volume of water balance components between pre- and post-relicensing operations could partially be explained by the precipitation difference in pre- and post-relicensing years (Figure 12).  During the pre-relicensing year, total output exceeded total input by approximately 21 billion cfd.  The post relicensing total output, on the other hand, exceeded total input by approximately 3.6 billion cfd.  Total ET was similar for both pre- and post relicensing operations.  The significant difference in pre-relicensing output to input was accounted as δ values and was about 17 billion cfd greater than the post-relicensing reservoir flux.

The difference in pre- and post-relicensing δ values implies that the management of the impoundments has altered the Muskegon River’s water balance.  Of the post-relicensing δ value of 2.78 billion cfd, 2.94 billion cfd was calculated to have entered the impoundments from tributaries and the total GW component was calculated to have been –0.17 billion cfd.  In contrast, during the pre-relicensing operations, of the δ value of 20.06 billion cfd, 5.78 billion cfd was calculated to have entered the impoundments from tributaries and the total GW component was calculated to have been 14.28 billion cfd.  These results suggest that, in addition to the greater tributary inflows during the pre-relicensing year, the GW component of the δ values during the pre-relicensing year 1991 was the primary difference in the total water balance.

January and February Flow Duration and Water Balance

The Hardy impoundment is drawn down (see Methods) during January and February of each year.  The pre- and post-relicensing input flows were similar at all probabilities, but the pre-relicensing output flows exceeded the post-relicensing output flows for all exceedance probabilities (Table 7, Figure 14).  The difference in Q50 flows for the pre-relicensing input and output was 34.20%.  Post-relicensing input and output flows were very close at Q5 flows and diverged at increasingly lower flows.  The difference in Q50 post-relicensing input and output flows was 16.92%.  The differences in shape of the pre- and post-relicensing input and output suggests that the post-relicensing flows were managed carefully at high flows, but as the flows became lower, matching of output to input was more difficult.  Alternatively, pre-relicensing input and output for January and February were consistently different by 28% or greater.

The water balance for January and February during the pre-relicensing year 1991 and post-relicensing year 2000 offered a more detailed perspective on the components that comprise the Muskegon River within the impoundments (Table 6, Figures 15 and 16).  During the pre-relicensing year, output exceeded input by 4.8 billion cfd.  The post relicensing output, on the other hand, exceeded total input by approximately 1.3 billion cfd.  January and February ET was similar for both pre- and post relicensing operations.  The significant difference in pre-relicensing output to input was accounted as δ values and was 3.5 billion cfd greater than the post-relicensing reservoir flux.

Both pre- and post-relicensing δ values included a dS/dt value of 1.3 billion cfd.  Of the pre-relicensing δ value of 4.76 billion cfd, 0.81 billion cfd was calculated to have entered the impoundments from tributaries and the total GW component was calculated to have been 5.25 billion cfd.  In contrast, during the post-relicensing operations, of the δ value of 1.28 billion cfd, 0.50 billion cfd was calculated to have entered the impoundments from tributaries and the total GW component was calculated to have been 2.08 billion cfd.  These results suggest that the GW component of the δ values during the pre-relicensing year 1991 and post-relicensing year 2000 was the primary difference in the January and February water balance.

Normalized to output, January and February water balance components reinforced the abovementioned findings (Table 8, Figures 17 and 18).  In both the pre-relicensing year 1991 input was 67% of output and in the post-relicensing year 2000, input was 82% of output.  ET was 0% of output for both pre- and post-relicensing years.  δ values represented the difference in input and output for pre- and post-relicensing years and were 33% and 17% of output, respectively.  Considering that the Hardy Impoundment was drawndown during this time period, the significant change in δ values were comprised primarily of the influx of GW into the impoundments suggesting that the reduced reservoir level, especially at pre-relicensing flows, at Hardy allowed greater GW to enter the system.

March and April Flow Duration and Water Balance

The Hardy impoundment is refilled (see Methods) during March and April of each year.  The pre-relicensing flows exceeded the post-relicensing flows for all exceedance probabilities (Table 9, Figure 19).  The difference in Q50 flows for the pre-relicensing input and output was 14.31%.  Post-relicensing input was quite similar to post-relicensing output for all.  The difference in Q50 flows for the pre-relicensing input and output was –3.70%.  The differences in shape of the pre- and post-relicensing input and output suggests that the post-relicensing flows were managed carefully.  Alternatively, pre-relicensing input and output for March and April appeared to be more difficult to manage at the extreme low flows and were different by as much as 35% or greater.

The water balance for March and April during the pre-relicensing year 1991 and post-relicensing year 2000 offered a more detailed perspective on the components that comprise the Muskegon River within the impoundments (Table 6, Figures 20 and 21).  During the pre-relicensing year, output exceeded input by 2.9 billion cfd.  The post relicensing output, on the other hand, exceeded total input by approximately 0.2 billion cfd.  March and April ET was similar for both pre- and post relicensing operations.  The significant difference in pre-relicensing input to output was calculated to be δ values and was 2.7 billion cfd greater than the post-relicensing reservoir flux.

Both pre- and post-relicensing δ values included a dS/dt value of –1.3 billion cfd.  Of the pre-relicensing δ value of 2.76 billion cfd, 1.35 billion cfd was calculated to have entered the impoundments from tributaries and the total GW component was calculated to have been 0.10 billion cfd.  In contrast, during the post-relicensing operations, of the δ value of 0.05 billion cfd, 0.60 billion cfd was calculated to have entered the impoundments from tributaries and the total GW component was calculated to have been –1.85 billion cfd.  These results suggest that the tributary and GW components of the δ values during the pre-relicensing year 1991 and post-relicensing year 2000 were the primary difference in the March and April water balance.

Normalized to output, March and April water balance components were quite different (Table 8, Figures 22 and 23).  In the pre-relicensing year 1991, input was 87% of output and during the post-relicensing year 2000, input was 100% of output.  ET was –1% of output for both pre- and post-relicensing years.  δ values represented the difference in input and output for pre- and post-relicensing years and were 13% and –1% of output, respectively.  Considering that the Hardy Impoundment was refilled during this time period, the significant change in δ values was comprised primarily of the tributaries and GW entering the impoundments during the pre-relicensing year 1991.  This implies that the reservoir flux at the Hardy impoundment allowed greater GW to enter the system.

May through August Flow Duration and Water Balance

The pre-relicensing flows exceeded the post-relicensing flows for all exceedance probabilities from May through August (Table 10, Figure 24).  The pre-relicensing output exceeded the pre-relicensing input for all flows.  The difference in Q50 flows for the pre-relicensing input and output was 27.1%.  Post-relicensing output exceeded the post-relicensing input for all flows as well.  The difference in Q50 flows for the post-relicensing input and output was 4.7%.  The shape of the pre- relicensing input and output suggests that flows were not managed closely as output exceeded input by 18.6% or greater.  Alternatively, post-relicensing input and output for May through August were managed more closely at all flows as output exceeded input by no greater than 6.4%.

The water balance for May through August during the pre-relicensing year 1991 and post-relicensing year 2000 offered a more detailed perspective on the components that comprise the Muskegon River within the impoundments (Table 6, Figures 25 and 26).  During the pre-relicensing year, output exceeded input by 5.8 billion cfd.  The post relicensing output, on the other hand, exceeded input by approximately 1.1 billion cfd.  May through August ET was similar for both pre- and post relicensing operations.  The difference in pre-relicensing output to input was accounted as δ values and was 4.4 billion cfd greater than post-relicensing reservoir flux.

Of the pre-relicensing δ value of 5.26 billion cfd, 1.51 billion cfd was calculated to have entered the impoundments from tributaries and the GW component was calculated to have been 3.75 billion cfd.  In contrast, during the post-relicensing operations, of the δ value of 0.65 billion cfd, 1.08 billion cfd was calculated to have entered the impoundments from tributaries and the total GW component was calculated to have been –0.43 billion cfd.  These results suggest that the GW component of the δ values during the pre-relicensing year 1991 and post-relicensing year 2000 was the primary difference in the May through August water balance.

Normalized to output, May through August water balance components were quite dissimilar (Table 8, Figures 27 and 28).  In the pre-relicensing year 1991, input was 77% of output.  In the post-relicensing year 2000, input was 96% of output, respectively.  ET was –2% and –3% of output for the pre- and post-relicensing years, respectively.  δ values represented the difference in input and output for pre- and post-relicensing years and were 20% and 1% of output, respectively.  Considering that the operations of the hydroelectric facilities were considerably different during pre –and post-relicensed years, the overall difference in water balance components seems reasonable.  The substantial differences suggests that operations of the hydroelectric facilities, during times of normal operations of the Hardy Impoundment, could have had significant impact on the Muskegon River system.

September through December Flow Duration and Water Balance

The pre-relicensing flows exceeded the post-relicensing flows for all exceedance probabilities from September through December (Table 11, Figure 29).  The pre-relicensing output exceeded the pre-relicensing input for all flows.  The difference in Q50 flows for the pre-relicensing input and output was 19.9%.  Post-relicensing output exceeded the post-relicensing input for all flows.  The difference in Q50 flows for the post-relicensing input and output was 3.4%.  The shape of the pre- relicensing input and output suggests that flows were not managed closely during September through December.  Alternatively, post-relicensing input and output differences were consistently less than pre-relicensing differences.  Post-relicensing flows were managed more carefully at all flows but differences did become greater at the lowest flows.

The water balance for September through December during the pre-relicensing year 1991 and post-relicensing year 2000 offered a more detailed perspective on the components that comprise the Muskegon River within the impoundments (Table 6, Figures 30 and 31).  During the pre-relicensing year, output exceeded input by 7.5 billion cfd.  The post relicensing output, on the other hand, exceeded input by approximately 1.0 billion cfd.  September through December ET was similar for both pre- and post relicensing operations.  The difference in pre-relicensing output to input was calculated to be δ values and was 6.6 billion cfd greater than post-relicensing reservoir flux.

Of the pre-relicensing δ value of 7.28 billion cfd, 2.10 billion cfd was calculated to have entered the impoundments from tributaries and the total GW component was calculated to have been 5.18 billion cfd.  In contrast, during the post-relicensing operations, of the δ value of 0.80 billion cfd, 0.78 billion cfd was calculated to have entered the impoundments from tributaries and the total GW component was calculated to have been 0.03 billion cfd.  These results suggest that the tributary input and GW components of the δ values during the pre-relicensing year 1991 and post-relicensing year 2000 were the primary difference in the September through December water balance.

Normalized to output, September through December water balance components were, again, relatively dissimilar (Table 8, Figures 32 and 33).  In the pre-relicensing year 1991, input was 78% of output.  In the post-relicensing year 2000, input was 93% of output, respectively.  ET was -1% of output for pre-relicensing operations and –2% of output for post-relicensing operations.  δ values represented the difference in input and output for pre- and post-relicensing years and were 21% and 6% of output, respectively.  Considering that the operations of the hydroelectric facilities were considerably different during pre –and post-relicensed years, it appears that the hydroelectric facilities were operated in a more consistent, run-of-river scheme during the post-relicensing year versus the pre-relicensing year.  Overall, it appears that the operations of the facilities had significant impact on the Muskegon River system during September through December.

Double Mass Analysis

In the double mass plot comparing the responsiveness of discharge to precipitation, pre- and post-relicensing discharge appeared to have a similar response to precipitation up to a cumulative precipitation of approximately 90 inches (Figure 34).  Above a cumulative precipitation of 90 inches, the pre-relicensing discharge became more responsive to cumulative precipitation.  The linear regression lines drawn through the curves suggest that post-relicensing cumulative discharge is less responsive to precipitation than pre-relicensing cumulative discharge.

            The double mass plot of cumulative input versus cumulative output for 1970 through 1973, 1974 through 1977, 1978 through 1981, 1982 through 1985, 1986 through 1989, pre- relicensing (1990 through 1993), and post-relicensing (1996 thorugh 1999, and 2000) hydroelectric operations at the Rogers, Hardy, and Croton facilities implied a trend towards a lower input to output ratio (Figure 35).  Post-relicensing operations have resulted in an output to input ratio closer to one-to-one.  The pre-relicensing operations (1990 through 1993) resulted in more output for given input.  The data for the additional pre-relicensing years, 1970 through 1989, suggested that the post-relicensing operations were not a result of climate induced operations.

            A scatter plot of input versus output for the pre- and post relicensing years with trend lines offered a perspective of the tendencies of the two operational strategies (Figure 36).  Post-relicensing operations exhibited a one-to-one input to output ratio, while pre-relicensing operations yielded more output for given input.

Energy and Erosion Potential

            Pre-relicensing stream power exceeded post-relicensing stream power implying that the peaking operations of the Muskegon River dams resulted in greater stream power than the run-of-river operations (Table 12).  Consideration must be given to the precipitation difference in pre- and post-relicensed years however (Figure 12).  The greatest difference in stream power occurred from September to December while the smallest difference occurred May through August.


DISCUSSION

Hydrologic Regime and Energy Impacts

            One of the goals of relicensing the Muskegon River dams was to help reduce the impacts the facilities have had on the river’s hydrologic regime.  To accomplish this goal, the FERC relicensed the facilities so that the output of the hydroelectric facilities (i.e. Croton Gauge) would have to approximately equal the input to the facilities (i.e. Big Rapids Gauge) plus the flow at the Little Muskegon Gauge (Consumers Power Company, 1998).  The result of this operational change has in fact altered flow regimes for the Muskegon River below the dams, but in unexpected as well as expected ways.

            During both both the pre- and post-relicensing periods , routing through Rogers, Hardy, and Croton dams resulted in a predictable hydrograph peak delay of two days  and little to no hydrograph attenuation, but instead hydrograph accretion (Figures 10 and 11).  This surprising result implies very significant groundwater loading between in the reach between Big Rapids and Croton.  During the Post-relicensing period, and following the agreement, the Dam outputs were managed to minimize this discrepancy. An inspection of the input and output flow duration curves indicate this management was effective at all exceedence frequencies (i.e., Q5 through Q95) (Figure 13). But, and presumably unexpectedly, the absolute dishrage from the impounded reach appears to have significantly declined as well.  The post-relicensing Q50 input was only 4.25% higher than the output, while the pre-relicensing Q50 input was 27.56% higher than the output (Table 5).

Pre- and post-relicensing water balance analyses offered another view of the impacts the change in operations of the facilities had on the individual components of the system (Tables 6 and 8).  Of particular interest was the impact the altered operations had on the δ values of the water balance equation.  The δ value for the entire pre-relicensing year was 20.06 billion cfd.  The δ value for the post-relicensing year was 2.78 billion cfd. The post-relicensing operations appear to have resulted in a more consistent  flow downstream of Croton, but also, on average, in lower flows measured at the Croton gauge.

Some portion of these lower flows could be explained by the lower rate of precipitation in the post-relicensing years (Figure 12), but, double mass and input output analyses suggest that lower precipitation cannot account for all of the observed decline (Figures 34-36)..  A  significant component of the water balance that appears to have been largely ignored in the relicensing of the facilities was ground water accrual.  The soils surrounding the impoundments are highly conductive, and discharge accretion rates clearly indicate substantial flux. However, the run-of-river scheme proposed did not appear to explicitly factor this into the flow regulation equation (Consumers Power, 1994a).

            Lower flows measured during post-relicensing operations could be a result of a reduced groundwater component in the water balance.  There is little data available for reservoir head during pre-relicensing operations, but it seems reasonable to assume that the Croton reservoir level was less stable during the pre-relicensing period because eof peaking operations of the dams.  The goal of the post-relicensing operations of Croton dam has been to maintain the level of the Croton reservoir at 722 feet above sea level, plus or minus one foot.  The result of this constant head could be a suppression of groundwater flux into and out of the reservoir through the exertion of a more consistent downward pressure within the reservoir (Dunne and Leopold, 1978).  The difference in pre- and post-relicensing operations on the head level of Croton reservoir should be further investigated as a potential cause of the observed reduction in flows through the impoundments.

That the relicensing of the three hydropower facilities on the lower river has impacted the flow regime by reducing short-term (daily) peaking oscillations is clear. Whether or not reduced overall flow was an expected (or  is now desirable) consequence of the licensing restrictions is less clear. Reduced flows downstream of the facilities necessarily affect the energy dissipation rate and capacity for the Muskegon River to do geomorphic work. The total power dissipation rate of the Muskegon River appears to have been substantially reduced in the post-relicensing phase of the facilities (Table 12).  Sediment accretion and channel aggradation appears to be a major feature of the lower river’s behavior in recent (Wiley, 2003; Richards, 2003).

What seems clearest at this point is that the lower Muskegon River, post-relicensing, has carried less water, on average, than during pre-relicensing operations (Table 6).  The river could be in the process of altering its slope to accommodate for the altered stream power.  The potential impact of the suppressed groundwatere accrual in the vicinity of the reservoirs certainly needs to be component considered in future modeling of scenarios involving Dam management.  The current FERC licenses for the three hydroelectric facilities on the Muskegon River will expire in 2034.  Additional flow data gathered throughout this period could offer a more accurate depiction of the impacts that the current run-of-river release strategy is having on the Muskegon River.

                                                                                                         

 

 

Limitations and Caveats

Several important data limitations, which could affect our observations and conclusions, must be made clear.  First, the lack of consistent gauging sites above and below the impounded segments meant that in this analysis we relyedheavily on gauge correlations using linear regression to construct hydrographic time-series (figures 6-8).  While every attempt was made at accuracy, there can be significant error in using such relationships to reconstruct flow histories. We believe, however, based on an inspection of the residuals that while significant estimation error is likely in some of the calculations, there is no directional bias in these calibrations so that absolute errors in estimated discharge should be randomly distributed with respect to over or under estimation. Another limitation of this study was the lack of water level information in Rogers, Hardy, and Croton impoundments.  This data, for both pre- and post-relicensing operations, if available, would offer a good perspective on how the operations of the facilities might be impacted groundwater flux into and out of the impoundments.  Finally, and perhaps most importantly, a longer data record is required to ultimately determine if the trends observed here are inherent in the current operation strategy or a transient result of regional hydrology and climate variation..  Pre-relicensing streamflow information for Evart, Little Muskegon at Morley, and Newaygo exists for 1970 through 1993.  Post-relicensing streamflow information for Big Rapids, Little Muskegon at Oak Grove, and Croton only exists from 1996 to present.  As more data is gathered for post-relicensing operations, a more accurate observation of the impact on the Muskegon River of these operations will be possible

Potential Ecological Impacts

            Any systematic alteration of the hydrologic regime and related changes in the geomorphic energies of the Muskegon River below Croton impoundment likely will have ecological implications.  The primary goals of the run-of-river operational scheme originally were to lower water temperatures on the lower Muskegon and maintain a more stable, oxygenated water quality.  Managing the river-flood plain interactions and reducing erosive scouring events was also a goal of the relicensing agreement (Consumers Power Company, 1994a).

            If the 1994 release strategy is changing rates of groundwater flux within the Muskegon River impoundments, the relicensing could be having a rather complex impact on the river’s temperature. We have not to date examined the thermal regime in a way that would be analogous to our hydrographic analysis here. However, if the groundwater component of the water balance continues to be reduced, river temperature could increase despite the reduction in reservoir residence times, because of the loss of relatively colder groundwater sources. In addition, there are a variety of concerns relating to the reduction in floodplain interaction on the lower Muskegon River that revolve around the river’s reduced power downstream of the hydroelectric facilities.  Significant power reduction will inevitably lead to the river’s reduced ability to transport sediment load.  The resultant channel instability could lead to further erosion and additional channel aggradation in the estuary zone of the river.

            The importance of the interaction between a river and its floodplain has only recently been acknowledged.  Previously thought to be a different ecosystem entirely, the floodplain is now considered an important contributor to river productivity, allowing it to obtain and carry needed nutrients for the organisms it supports (Allan, 1995).  As stated in the Multiple Projects Environmental Assessment for the Muskegon River projects (1994a), the operation scheme will “minimize the inundation of the downstream floodplain”.  A concern for the future management of the Muskegon River could be the impact the altered hydrologic regime will have on the river’s interaction with the floodplain.

 


 

Table 1.  Physical characteristics of the Rogers, Hardy, and Croton impoundments.

Impoundment Characteristics

Rogers

Hardy

Croton

Location (RM)

83.7

58.7

52.8

Drainage Area (mi2)

~1,746

~1,851

~2,224

Max. Surface Area (acres)

449

3,902

1,209

Max. Surface Level (ft)

861.6

822.5

722.0

Gross Storage Capacity (acre-ft)

4,678

134,973

21,932

Target Surface Level (ft)

861.3

822.0

722.0

Mean Depth (ft)

10.4

34.5

18.1

Normal Usable Storage Capacity (acre-ft) / Impoundment Fluctuation (ft)

300/0.7

5,007/1.3

1,181/1.0

Normal Impoundment Fluctuation (ft)

0.7

1.0

1.0

Drawdown (ft.) : dewatered acres (% surface area)

N/A

0.5 : 51 (1.3%)

1.0 : 57 (4.7%)

Average Annual Flow (cfs)[1]

1,344

1,460

1,871

Impoundment Retention Time (days) @ Avg. Annual Flow1

1.8

46.6

5.9

Flow exceeded 90% of time (cfs)1

650

650

1,000

Flow exceeded 10% of time (cfs)1

2,850

2,850

3,600

 

Table 2.  Operating characteristics of the Rogers, Hardy, and Croton impoundments.

Operating Characteristics

Rogers

Hardy

Croton

Operating Mode

Run-of-River

Ponding and Peaking from 8am-11pm

Re-regulated flow release

Min. Turbine Hydraulic Range (cfs) / Flow Duration Curve

300 / 99%

1,600 / 27%

300 / 99%

Max. Turbine Hydraulic Range (cfs) / Flow Duration Curve

2,420 / 15%

4,860 / 2.5%

3,700 / 10%

Subject to Winter Drawdown & Spring Refill

No

Yes

No

Max. Winter Drawdown (ft) / Drawdown Timing

N/A

12 / January & February

N/A

Drawdown Impacted Surface Area / Volume (acre-ft)

N/A

18%-25% / ~30,000

N/A

Operations During Winter Drawdown

Run-of-River

Winter Drawdown

Releases mean daily discharge from Hardy + Little Muskegon R. inflow

 


Table 3.  Land use characteristics of the Rogers, Hardy, and Croton impoundments.

Land Use

Rogers

Hardy

Croton

Project Acreage / Impoundment Acreage

2,200 / 449

5,000 / 3,902

2,200 / 1,209

Shoreline (mi)

~19.4

~50

~22.9

CP Owned

18%

99.6%

11.3%

Agricultural

1%-4%

1%-4%

1%-4%

Forested

76%

81%

79%

Residential

17%

-

17%

Open Gravel Pits

2.7%

-

-

 

Table 4.  USGS gauging station information for selected sites on the Muskegon and Little Muskegon Rivers.

Station Name

Site Number

Period of

Record

Latitude

Longitude

Catchment 

(sq. mi)

Muskegon River at Evart

4121500

1930-present

43.8992

85.2553

1,433

Muskegon River at Big Rapids

4121650

1999-present

43.6936

85.4675

1,751

Little Muskegon at Morley

4121900

1966-1996

43.5025

85.3425

121

Little Muskegon at Oak Grove

4121944

1995-present

43.4308

85.5956

345

Muskegon River near Croton

4121970

1995-present

43.4347

85.6653

2,313

Muskegon River at Newaygo

4122000

1930-1993

43.4222

85.8011

2,350

 

Table 5.  Flow duration percentiles and percent difference for pre-relicensing (1990-1993) input and output and post-relicensing (1996-2000) input and output for the Muskegon River.

 

5th Percentile

25th Percentile

50th Percentile

75th Percentile

95th Percentile

Pre-relicensing Input (cfs)

4,458

2,818

1,965

1,532

1,186

Pre-relicensing Output (cfs)

5,023

3,568

2,713

2,162

1,624

% Difference

11.25%

21.02%

27.56%

29.13%

26.94%

Post-relicensing Input (cfs)

3,404

2,085

1,513

1,163

834

Post-relicensing Output (cfs)

3,400

2,200

1,580

1,230

885

% Difference

-0.11%

5.23%

4.25%

5.43%

5.76%

 

 

 

 

 

 

 

 

 

Table 6.  Hydrograph flows for input, output, ET, and reservoir flux (δ) for 1991 and 2000 for the Muskegon River impoundments.

Date

Total Input

Total Output

Total ET

Total Res. Flux

 

(billion cfd)

(billion cfd)

(billion cfd)

(billion cfd)

1991 Total

80.39

101.40

-0.95

20.06

2000 Total

48.03

51.67

-0.85

2.78

Jan & Feb 1991

9.49

14.29

-0.04

4.76

Jan & Feb 2000

7.52

8.83

-0.04

1.28

Mar & Apr 1991

20.84

23.74

-0.14

2.76

Mar & Apr 2000

10.18

10.37

-0.13

0.05

May – Aug 1991

20.71

26.53

-0.56

5.26

May – Aug 2000

17.72

18.86

-0.49

0.65

Sept – Dec 1991

29.34

36.84

-0.22

7.28

Sept – Dec 2000

12.61

13.61

-0.19

0.80

 

Table 7.  Flow duration percentiles and percent difference for pre-relicensing (1990-1993) and post-relicensing (1996-2000) input and output for the Muskegon River in January and February.

 

5th

Percentile

25th

Percentile

50th

Percentile

75th

Percentile

95th

Percentile

Pre-relicensing Input (cfs)

2,614

2,035

1,837

1,578

1,246

Pre-relicensing Output (cfs)

3,668

3,025

2,791

2,519

2,126

% Difference

28.75%

32.74%

34.20%

37.35%

41.37%

Post-relicensing Input (cfs)

3,389

2,281

1,820

1,420

1,120

Post-relicensing Output (cfs)

3,480

2,620

2,190

1,710

1,380

% Difference

2.61%

12.92%

16.92%

16.93%

18.84%

 

Table 8.  Average hydrograph flows for input, output, ET, and reservoir flux (δ) normalized to output for 1991 (pre-relicensing) and 2000 (post-relicensing) for the Muskegon River impoundments.

Date

Average Input

Average Output

Average ET

Average Res. Flux

Jan & Feb 1991

67%

100%

0%

33%

Jan & Feb 2000

82%

100%

0%

17%

Mar & Apr 1991

87%

100%

-1%

13%

Mar & Apr 2000

100%

100%

-1%

-1%

May – Aug 1991

77%

100%

-2%

20%

May – Aug 2000

96%

100%

-3%

1%

Sept – Dec 1991

78%

100%

-1%

21%

Sept – Dec 2000

93%

100%

-2%

6%

 

 

 

Table 9.  Flow duration percentiles and percent difference for pre-relicensing (1990-1993) and post-relicensing (1996-2000) input and output for the Muskegon River in March and April.

 

5th

Percentile

25th

Percentile

50th

Percentile

75th

Percentile

95th

Percentile

Pre-relicensing Input (cfs)

5,759

4,300

3,228

2,411

1,601

Pre-relicensing Output (cfs)

5,628

4,800

3,768

2,861

2,466

% Difference

-2.34%

10.42%

14.31%

15.74%

35.09%

Post-relicensing Input (cfs)

4,662

3,157

2,468

1,913

1,515

Post-relicensing Output (cfs)

4,720

3,160

2,380

1,820

1,540

% Difference

1.23%

0.10%

-3.70%

-5.11%

1.65%

 

Table 10.  Flow duration percentiles and percent difference for pre-relicensing (1990-1993) and post-relicensing (1996-2000) input and output for the Muskegon River in May through August.

 

5th

Percentile

25th

Percentile

50th

Percentile

75th

Percentile

95th

Percentile

Pre-relicensing Input (cfs)

3,293

2,177

1,623

1,340

1,151

Pre-relicensing Output (cfs)

4,046

2,791

2,225

1,890

1,540

% Difference

18.62%

22.00%

27.07%

29.12%

25.27%

Post-relicensing Input (cfs)

3,168

1,933

1,344

1,040

785

Post-relicensing Output (cfs)

3,370

1,980

1,410

1,110

838

% Difference

6.00%

2.37%

4.66%

6.31%

6.38%

 

Table 11.  Flow duration percentiles and percent difference for pre-relicensing (1990-1993) and post-relicensing (1996-2000) input and output for the Muskegon River in September through December.

 

5th

Percentile

25th

Percentile

50th

Percentile

75th

Percentile

95th

Percentile

Pre-relicensing Input (cfs)

4,048

2,847

2,018

1,770

1,204

Pre-relicensing Output (cfs)

4,923

3,660

2,519

2,323

1,670

% Difference

17.78%

22.22%

19.89%

23.82%

27.87%

Post-relicensing Input (cfs)

2,076

1,547

1,285

1,090

809

Post-relicensing Output (cfs)

2,170

1,590

1,330

1,120

872

% Difference

4.35%

2.69%

3.38%

2.68%

7.28%

 

Table 12.  Pre-relicensing (1991) and post-relicensing (2000) stream power and percent difference for the Muskegon River from the tailwaters of Croton Dam to Muskegon Lake.

Date

Pre-relicensing

Stream Power (MWh ft-1)

Post-relicensing

Stream Power (MWh ft-1)

% Difference

Stream Power (MWh ft-1)

Jan. & Feb.

2,735

1,641

39.99%

Mar. & Apr.

4,643

2,130

54.12%

May. – Aug.

10,804

7,565

29.98%

Sept. – Dec.

14,481

5,060

65.05%

 

 


Source: Muskegon River Watershed Assembly

 

Figure 1.  The Muskegon River watershed surrounding the Rogers, Hardy, and Croton dams.

 

Figure 2.  Linear representation of the inflows and outflows of the Rogers, Hardy, and Croton impoundments on the Muskegon River.

 

 

Figure 3.  The Rogers hydroelectric facility on the Muskegon River, Mecosta County.

 

 

Figure 4.  The Hardy hydroelectric facility on the Muskegon River, Newaygo County.

 

 

Figure 5.  The Croton hydroelectric facility on the Muskegon River, Newaygo County.

Figure 6.  Scatter plot and linear regression for correlation of measured stream flow at Evart gauging station to Big Rapids gauging station.

Figure 7.  Scatter plot and linear regression for correlation of measured stream flow at Little Muskegon at Morley gauging station to Little Muskegon at Oak Grove gauging station.

Figure 8.  Scatter plot and linear regression for correlation of measured stream flow at Newaygo gauging station to Croton gauging station.

Figure 9.  Water balance components for Muskegon River hydrologic equation.

 

 

 

 

 

Figure 10.  Fifteen-minute hydrograph for the year 1991.  Input flows determined using the Evart to Big Rapids correlation  and output flows determined using the Newaygo to Croton correlation.

Figure 11.  Fifteen-minute hydrograph for 2000.  Input flows measured at the Big Rapids gauging station and output flows measured at the Croton gauging station.

Figure 12.  Precipitation exceedance probabilities for pre-relicensing (1990-1993) and post-relicensing (1996-2000) years.

Figure 13.  Flow duration curve for pre-relicensing (1990-1993) and post-relicensing (1996-2000) input and output for the Muskegon River.

Figure 14.  Flow duration curve for pre-relicensing (1990-1993) and post-relicensing (1996-2000) input and output in January and February for the Muskegon River.

Figure 15.  Hydrograph for the Muskegon River input, output, ET, and reservoir flux (δ), including dS/dt, small tributary inputs, and GW, for January and February 1991.

Figure 16.  Hydrograph for the Muskegon River input, output, ET, and reservoir flux (δ), including dS/dt, small tributary inputs, and GW, for January and February 2000.

Figure 17.  Hydrograph for the Muskegon River input, output, ET, and reservoir flux (δ), including dS/dt, small tributary inputs, and GW, normalized to output for January and February 1991.

Figure 18.  Hydrograph for the Muskegon River input, output, ET, and reservoir flux (δ), including dS/dt, small tributary inputs, and GW, normalized to output for January and February 2000.

Figure 19.  Flow duration curve for pre-relicensing (1990-1993) and post-relicensing (1996-2000) input and output in March and April for the Muskegon River.

Figure 20.  Hydrograph for the Muskegon River input, output, ET, and reservoir flux (δ), including dS/dt, small tributary inputs, and GW, for March and April 1991.

Figure 21.  Hydrograph for the Muskegon River input, output, ET, and reservoir flux (δ), including dS/dt, small tributary inputs, and GW, for March and April 2000.

Figure 22.  Hydrograph for the Muskegon River input, output, ET, and reservoir flux (δ), including dS/dt, small tributary inputs, and GW, normalized to output for March and April 1991.

Figure 23.  Hydrograph for the Muskegon River input, output, ET, and reservoir flux (δ), including dS/dt, small tributary inputs, and GW, normalized to output for March and April 2000.

Figure 24.  Flow duration curve for pre-relicensing (1990-1993) and post-relicensing (1996-2000) input and output in May through August for the Muskegon River.

Figure 25.  Hydrograph for the Muskegon River input, output, ET, and reservoir flux (δ), including dS/dt, small tributary inputs, and GW, for May through August 1991.

Figure 26.  Hydrograph for the Muskegon River input, output, ET, and reservoir flux (δ), including dS/dt, small tributary inputs, and GW, for May through August 2000.

Figure 27.  Hydrograph for the Muskegon River input, output, ET, and reservoir flux (δ), including dS/dt, small tributary inputs, and GW, normalized to output for May through August 1991.

Figure 28.  Hydrograph for the Muskegon River input, output, ET, and reservoir flux (δ), including dS/dt, small tributary inputs, and GW, normalized to output for May through August 2000.

Figure 29.  Flow duration curve for pre-relicensing (1990-1993) and post-relicensing (1996-2000) input and output in September through December for the Muskegon River.

Figure 30.  Hydrograph for the Muskegon River input, output, ET, and reservoir flux (δ), including dS/dt, small tributary inputs, and GW, for September through December 1991.

Figure 31.  Hydrograph for the Muskegon River input, output, ET, and reservoir flux (δ), including dS/dt, small tributary inputs, and GW, for September through December 2000.

Figure 32.  Hydrograph for the Muskegon River input, output, ET, and reservoir flux (δ), including dS/dt, small tributary inputs, and GW, normalized to output for September through December 1991.

Figure 33.  Hydrograph for the Muskegon River input, output, ET, and reservoir flux (δ), including dS/dt, small tributary inputs, and GW, normalized to output for September through December 2000.

Figure 34.  Double mass plot for cumulative precipitation versus cumulative output for pre-relicensed (1990-1993) and post-relicensed (1996-2000) data.

‘86-‘89

 

‘78-‘81

 

‘74-‘77

 

‘82-‘85

 

‘70-‘73

 

‘96-‘99

 

‘00

 

‘90-‘93

 

Figure 35.  Double mass plot for cumulative input versus cumulative output for 1970 – 1973, 1974 – 1977, 1978 – 1981, 1982 – 1985, 1986 – 1989, pre-relicensing (1990 – 1993) and post-relicensed data (1996-2000).

Figure 36.  Scatter plot of input versus output for pre-relicensing (1990-1993) and post-relicensing (1996-2000) operations.


LITERATURE CITED

Alexander, J.  1999a.  A shadow of its former self.  Part 1: The Muskegon River Unnatural Wonder.  The Muskegon Chronicle.  September 12, 1999.

 

Alexander, J.  1999b.  Hydroelectric dams kill thousands of sport fish annually.  Part 1: The Muskegon River Unnatural Wonder.  The Muskegon Chronicle.  September 12, 1999.

 

Alexander, J.  1999c.  A river journal.  Part 1: The Muskegon River Unnatural Wonder.  The Muskegon Chronicle.  September 12, 1999.

 

Allan, J.D.  1995.  Stream ecology: structure and function of running waters.  Kluwer Academic Publishers, Dordrecht, 388 pp.

 

Boutin, C.  2000.  Great lakes water levels.  NOAA Great Lakes Environmental Research Laboratory, Ann Arbor, 2 pp.

 

Chin, A., Harris, D.L., Trice, T.H., and J.L. Given.  2002.  Adjustment of stream channel capacity following dam closure, Yegua Creek, Texas.  Journal of the American Water Resources Association, 38(6):1521-1531.

 

Consumers Power Company.  1998.  Manual re-regulation testing plan: Croton Dam, FERC No. 2468.  July 10, 1998.

 

Consumers Power Company.  1994a.  Final Multiple Project Environmental Assessment for New Hydropower Licenses.  Muskegon River Projects.  FERC Nos. 2451-004 (Rogers), 2452-007 (Hardy), and 2469-003 (Croton).  January 5, 1994.

 

Consumers Power Company.  1994b.  Order Issuing New License.  Rogers Hydroelectric Project.  FERC No. 2451-004.  July 15, 1994.

 

Consumers Power Company.  1994c.  Order Issuing New License.  Hardy Hydroelectric Project.  FERC No. 2452-007.  July 15, 1994.

 

Consumers Power Company.  1994d.  Order Issuing New License.  Croton Hydroelectric Project.  FERC No. 2468-003.  July 15, 1994.

 

Dunne, T. and L.B. Leopold.  1978.  Water in environmental planning.  W.H. Freeman and Company, New York, 818 pp.

 

Dynesius, M. and C. Nilsson.  1994.  Fragmentation and flow regulation of river systems in the northern third of the world.  Science, 266:753-762.

 

Geis, J.  1982.  Hydropower’s unexpected side effects from “Hydropower in Minnesota’s Future”.  The Minnesota Volunteer.  September-October, 1982.

 


 

Graf, W.L.  1992.  Science, public policy, and western american rivers.  Transactions of the Institute of British Geographers, 17:5-19.

 

Grant, G.E., Schmidt, J.C., and S.L. Lewis.  2003.  A geological framework for interpreting downstream effects of dams on rivers. in J.E. O'Connor and G.E. Grant eds.  A Peculiar River. Water Science and Application 7. American Geophysical Union, 209-225.

 

Knighton, A.D.  1998.  Fluvial Forms and Processes: A New Perspective.  Arnold, London, 383 pp.

 

Linsley, R., M. Kohler, and J. Paulhus.  1982.  Hydrology for Engineers, McGraw Hill, 508 pp.

 

O’Neal, R.  1997.  Muskegon River Watershed Assessment.  Michigan Department of Natural Resources Special Report No. 19, 187 pp.

 

Richards, P.L.  2003.  Personal communication.

 

Seelbach, P.  1999.  Quote from: “A shadow of its former self.”  Part 1: The Muskegon River Unnatural Wonder.  The Muskegon Chronicle.  September 12, 1999.

 

Sluka, S.  2002.  Interview with Croton hydroelectric facilities manager.  September 13, 2002.

 

Snyder, R.L.  2000.  Monthly Reference Evapotranspiration (ETo) Calculator.  Regents of the University of California; website: http://lawr.ucdavis.edu/coopextn/biometeorology/evapotranspiration/PMmonXLS/PMmon.xls (May 7, 2003).

 

Wiley, M.J.  2003.  Personal communication.



[1] As calculated by CP using a 20-year (1967 to 1986) period of record collected at USGS gage #04121500 (Rogers & Hardy) and USGS gage #04122000 (Croton)