EX RPT & DWGS - 26149 2010 Principal Spillway Discharge ScourRED.pdf

Dam Undermining Investigation Report

Shiawassee Flats Dam - Radial Gates Shiawassee River State Game Area

St. Charles, Michigan

Prepared for the

Michigan Department of Natural Resources

AND THE

Michigan Department of Management and Budget

FILE NO. 751/08274.RWG - INDEX NO. 86930

NOVEMBER 25, 2009

BY

Williams $ Works engineers . planners . surveyors tradition of service

616.224.1500 phone . 800.224.1590 toll free . 616 224 1501 facsimile 549 Ottawa Avenue NW . Grand Rapids. Ml 49503

RECEIVED

JUN fl 7 2011) environmental quality

LAND & WATER MANAGEMENT DIV

Shiawassee Flats Dam - Radial Gates

Dam Undermining Investigation Report

1. Introduction

The Shiawassee River State Game Area is owned and operated by the Michigan Department of Natural Resources (MDNR) Wildlife Management Division. The State Game Area (SGA) includes an earthen dam which separates the Shiawassee River flats from the Bad River. This dam is known as the Shiawassee Flats Dam, Dam ID #1819. The dam’s primary water level (stage) control structure is a pair of radial gates, mounted integral to a small bridge, which control water movement from the Shiawassee River flats to the Bad River. The radial gates structure and its inlet/outlet channels are critical components of the State Game Area seasonal flooding operations

— controlling the flooding of the SGA’s Waterfowl Hunting Zones and wildlife management areas southeast of the Bad River.

2. Scope of Investigation

In May of 2009, Williams &. Works submitted an Erosion Control Measures Preliminary Design study of the Shiawassee Flats Dam to the Michigan Department of Natural Resources (MDNR) and the Michigan Department of Management & Budget (MDMB). During the course of that study concerns were raised about possible breaching of the inlet cutoff wall. Specifically, the bottom of the inlet scour hole was roughly equal to the bottom of the inlet cutoff wall. Additionally, some undermining was observed at the edge of the outlet spillway.

In August of 2009, the MDMB/MDNR contracted with Williams &. Works to perform a more detailed investigation of the potential undermining of the dam, consisting of:

^ Cofferdaming and dewatering of the Radial Gates structure and its outlet spillway > Concrete coring of the deck and spillway to investigate for voids > Probing of the observed outlet spillway undermining to characterize the soils loss > Diver inspection of the exposed inlet cutoff wall for damage and/or underflow > Remediation of voids as discovered using pressure grouting techniques > Study and report on the conditions of the Radial Gates structure and its outlet spillway

Following is Williams & Works report of the work performed and the conditions observed.

3. Outlet Spillway Investigation

In order for the work to be successful, the Shiawassee and Bad River levels had to fall below 3 ft on the spillway deck. They were observed for a suitable low water condition.

On August 26, 2009, the river levels dropped to 2.8 ft. above the spillway floor; low enough to begin the work. The initial cofferdam extended across the entire outlet edge of the spillway, and both radial gates were closed. However, during 2 days of cofferdam construction and pumping the subcontractor was unable to seal and dewater the work space. Since the spillway was constructed as a flat slab without cutoff walls, water was flowing freely under the slab, and up through the concrete joint where the spillway joined the radial gates structure. The spillway had apparently shifted downstream slightly, leaving a 2 in. to 3 in. wide gap which could not be sealed. Investigation of the depth of the void below the concrete floor varied as follows (also see Figure 1):

Southwestern corner: 16 in. deep Middle of southwest gate: 12 in. deep

Center pier: 5 in. deep Middle of northeast gate: 0 in. deep

Northeastern corner: 6 in. deep

Due to the inability to dewater the spillway area, and the poor visibility through the water, coring could not be performed successfully. Rain early on August 28th caused the river level to rise, interrupting further investigation until September 2nd.

On September 2nd, the diver conducted a probe investigation of the void below the spillway outlet edge. He found that the undercut measurements back up under the spillway, measured from the spillway edge, were as follow (also see Figure 1):

Southwestern corner: 5.1 ft. long Middle of southwest gate: 2.7 ft. long

Center pier: 2.5 ft. long Middle of northeast gate: 3.0 ft. long

Northeastern corner: 3.3 ft. long

Based on the adjacent clay slopes, we estimate that the undercut at the spillway outer edge varies from about 2 ft. deep to as much as 6 ft. deep. Based on the resulting cross-sectional areas of the undermining, the void volume may be as high as 600 cu. ft. (22.25 cu. yds.).

This also supports the belief that significant undermining at the southwest corner of the spillway has allowed it to settle slightly; about 2 inches.

It was not possible to grout the undercut zone within the authorized scope, since it would have required installation of water-tight sheeting along the outlet edge of the spillway.

4. Southwest Radial Gate Deck Coring Investigation

In order for the investigation to be successful, the cofferdam was relocated to the discharge of the southwest radial gate on September 2nd. Specifically, between the center pier and south abutment just inside the leaking spillway joint. Dewatering was successful and four

(4) cores were made; two (2) near the inlet side and two (2) near the outlet side on a staggered pattern. All of these cores were located between the radial gates inlet and outlet cutoff walls. Each core consisted of 17 to 20 in. thick concrete with reinforcing steel, poured on a stone base (apparently 6A angular stone), with no void space below the concrete. The core near the center pier of the Radial Gates (RG) inlet exhibited some seepage flow, reflecting a weak connection with the high water level upstream of the closed radial gate. Each core was plugged with a combination of concrete, burlap and quick-set hydraulic cement.

It was concluded that no significant voids existed below the southwest RG deck.

5. Northeast Radial Gate Deck Coring Investigation

Following completion of the southwest deck coring, the cofferdam was relocated to the discharge of the northeast radial gate on September 3rd. Specifically, between the center pier and north abutment just inside the leaking spillway joint. Dewatering was successful and four (4) cores were made; two (2) near the inlet side and two (2) near the outlet side on a staggered pattern. All of these cores were located between the radial gates inlet and outlet cutoff walls. Each core consisted of 18 in. to 22 in. thick concrete with reinforcing steel, poured on a stone base (apparently 6A angular stone), with no void space below the concrete. The two cores near the RG inlet exhibited seepage water flows, reflecting weak connections with the high water level upstream of the closed radial gate. Each core was plugged with a combination of concrete, burlap and quick-set hydraulic cement.

It was concluded that no significant voids existed below the northeast RG deck.

6. Radial Gates Inlet Cutoff Wall Investigation

During the initial dewatering work, the diver investigated the RG inlet cutoff wall. Prior surveying indicated that the bottom of the inlet scour hole may have breached the inlet cutoff wall. The diver’s investigation found:

o Five (5) feet of exposed sheeting at the center pier o A solid soil mass covering the bottom edge of the sheeting, and o The sides of the scour hole smoothly tapered up to the rip-rapped dikes.

It was concluded that the inlet cutoff wall had not been breached. Emergency backfill to within 1 ft. of the deck was done with layers of MDOT Plain and MDOT Large rip-rap.

7. Summary of Findings and Conclusions

Outlet Spillway & Cutoff Wall The radial gates spillway was constructed as a flat slab without cutoff walls. The spillway has apparently shifted downstream slightly, leaving a 2 in. to 3 in. wide gap where the spillway joins the radial gates structure. Significant undermining at the southwest corner of the spillway has allowed it to settle slightly; about 2 inches.

Water does flow under the slab, passing freely through the gap at the concrete joint. The radial gates outlet cutoff wall prevents the water from traveling under the floor of the gates.

Based on the approximate cross-sectional areas of the undermining, the void volume may be as high as 600 cu. ft. (22.25 cu. yds.). Future repair work should include sheeting and grouting of the void, or complete replacement of the spillway, as funding permits.

Radial Gates Deck Eight cores of the deck found concrete with reinforcing steel 17 to 22 in. thick, poured on a stone base (apparently 6A angular stone), with no void space below the concrete. Cores near the radial gates inlet exhibited a small water flow, reflecting a weak connection through the stone with the high water level upstream of the closed radial gates.

No significant voids exist below the radial gates deck.

Inlet Cutoff Wall Although the diver’s investigation found five (5) feet of exposed sheeting in front of the center pier, it was concluded that the inlet cutoff wall had not been breached. Emergency stone backfill was installed in the scour hole on September 8, 2009 to within 1 ft. of the RG deck elevation. A bottom layer of MDOT Plain rip-rap was followed by a top layer of MDOT Large rip-rap.

5, SO

O xn P.

\P ro

*s h\

8^

BO

uD

M B

\2 O

8l 0O lD \d

»g \F is uf tl- U nc «T

W n9

Jn m

'.i 9a

'.M Xi

.c Jw

9, XR

E:j : [208

0 BM

Si

CONCRETE CORE DATA*

CORE #

CONCRETE

THICKNESS

(INCHES)

1 19 2 19 3 17 4 20 5 18 6 18 7 22 8 20

Erosion Control Measures Preliminary Design Report

Shiawassee Flats Dam - Radial Gates Shiawassee River State Game Area

St. Charles, Michigan

Prepared for the

Michigan Department of Natural Resources and the

Michigan Department of Management and Budget

FILE NO. 751/O8274.RWG - INDEX NO. 86930

APRIL 6, 2009

BY

Williams $ Works engineers . planners . surveyors ^^a tradition of service

616.224.1500 phone 800.224.1590 toll free . 616.224.1501 facsimile

549 Ottawa Avenue NW . Grand Rapids, Ml 49503 received JUN o 7 2010

ENVIRGNMENTAL quality land & WATER MANAGEMENT DIV

Table of Contents

1. Introduction

2. Scope of Study

3. Project Location

4. Stream Names, Drainage Area & Discharge Rates

5. Radial Gates Existing Conditions

5.1. Structure Description/Observations

5.2. Inlet Description/Observations

5.3. Outlet Description/Observations

5.4. Bathymetry/Topography

5.5. Soil Conditions

5.6. FHWA Level One Scour Analysis

5.7. Summary of Erosion/Scour Impacts Observed

6. Radial Gates Hydraulics

6.1. Normal Operating Conditions

6.2. Flood Flow Operating Conditions................................................................. -—8

6.3. Discharge Equation Selection

6.4. Design Hydraulic Conditions Selection

7. Erosion/Scour Control Measures - Basis of Design

7.1. Inlet Channel Hydraulic Conditions Selected

7.2. Hydraulic Jump Stilling Basin Conditions Selected

7.3. Outlet Channel Hydraulic Conditions Selected

8. Erosion/Scour Control Measures - Alternatives Evaluation

8.1. Inlet Channel Alternatives

8.2. Spillway & Outlet Expansion Alternatives

8.3. Outlet Channel Alternatives

9. Erosion/Scour Control Measures - Recommendations

9.1. Emergency Erosion Control Measures

9.2. Long Term Erosion Control Measures and Improvements

9.3. Long Term Erosion Control Measures, Improvements and Capacity Enhancements

9.4. Construction Work Staging

Appendices

Appendix 1 - Existing Conditions Photos

Appendix 2 - Existing Topographic/Bathymetric Plans

Appendix 3 - Soil Sampling Particle Size Distribution

Appendix 4 - Level One Scour Analysis

Appendix 5 - Design Hydraulic Conditions Selection

Appendix 6 - Alternatives Cost Estimates

Page ii

Executive Summary

The Shiawassee River State Game Area is owned and operated by the Michigan Department of Natural Resources (MDNR) Wildlife Management Division. The State Game Area (SGA) includes an earthen dam which separates the Shiawassee River flats from the Bad River. This dam is known as the Shiawassee Flats Dam, Dam ID #1819. The Radial Gates flow control structure is located at the confluence of the Shiawassee River and the Bad River, and it controls flow from the Shiawassee River flats into the Bad River.

In March of 2008, the Dam Safety Program of the Michigan Department of Environmental Quality (MDEQ), Land and Water Management Division (LWMD), conducted an inspection of the Shiawassee Flats Dam. In his subsequent report, the inspector recommended that the MDNR:

Further investigate the erosion problem on the stream banks just downstream of the radial gates structure by December 31, 2008. On the basis of this study, install appropriate measures by December 31, 2009.

In November of 2008, the MDMB hired Williams & Works on behalf of the MDNR to perform the recommended investigation of the erosion. Williams &. Works (W&W) was also contracted to design, permit, bid and oversee construction of the selected erosion control measures.

The original scope of this project included repair of the erosion damage to the Radial Gates Outlet Channel banks pursuant to the MDEQ Dam Safety Inspection Report. The project budget was approximately $250,000. However, during the course of the Preliminary Design study, considerably more submerged damage was identified in the form of inlet and outlet scour holes. The inlet scour hole threatens to undermine the Radial Gates structure and cause dam failure. Currently, the bridge and Radial Gates structure appear to be stable, but the erosion/scour of the inlet and outlet channels must be corrected soon.

In addition, review of the past operation and the current hydraulic conditions indicate that the Radial Gates structure was not originally designed to operate alone. The decommissioned ‘Seven Tubes’ structure provided additional flow capacity and level control that was an important component of the dam’s hydraulic system. Given this background, the radial gates hydraulics were evaluated, hydraulic bases of design were selected, erosion repair/control alternatives were developed and evaluated, and three main strategies were identified to improve the situation, as follows:

1. Emergency Erosion Control Measures to minimize further scour/erosion, stabilize the Radial Gates structure, and avoid a complete failure of the Shiawassee Flats Dam,

2. Long Term Erosion Control Measures and Improvements to the Radial Gates to repair the existing damage and better protect it from the increased velocities and flows, or

3. Long Term Erosion Control Measures, Improvements along with Capacity Enhancements to return flow capacity and level control that was lost when the ‘Seven Tubes’ structure was decommissioned.

Page iii

Emergency Erosion Control Measures

The recommended Emergency Measures rely on the operational controls and emergency maintenance work identified in the ‘No Action’ Alternatives, I.O., S.O., E.O., and 0.0. In overview, this work consists of underwater inspection of the structure, rock filling of the scour holes, and repairing the channel banks in so far as possible without draining the inlet or outlet channels. The first year cost of this work is estimated at $210,000, plus annualized maintenance costs of $32,000 per year thereafter.

Emergency operational controls include placing specific limits on gate opening height in order to control the entry and discharge velocities, thereby minimizing additional scour and erosion. Generally these controls seek to limit the inlet, spillway and outlet channel average velocities to less than 5 fps. In some instances the operator will find the recommended gate opening limit to be very restrictive and inadequate to quickly drain the stored water and/or release flood flows in the desired timeframe. Since there are a wide array of headwater, tailwater and gate positions to be considered, tables of conditions and matching gate positions would be prepared for the operator’s guidance.

This combination of alternatives minimizes the immediate cost of protecting the radial gates from being undermined, and reduces the ongoing scour and erosion. However, the annual maintenance costs would be high and deviation from the gate operating limits plus annual flood flows through the open gates will still produce considerable erosion. Also, the loose-rock filling of scour holes now will increase the cost of future long term repairs. Most importantly, this strategy does not completely secure the structure from settlement and possible future collapse, and should not be considered a durable, long term remedy. This strategy should only be implemented if the current budget constraints are inflexible, and the long term control measures will not be implemented.

Long Term Erosion Control Measures and Improvements

The recommended Long Term Measures and Improvements include the following alternatives:

1.3. - Inlet Channel Repair/Restore, Concrete Inlet Apron and Bottom/Bank Armor (12” rock)

S.2. - Hydraulic Jump Stilling Basin with Concrete Side Walls (SAF class trapezoidal basin)

E.3. - Armored Outlet Expansion (12” rock) with Concrete Wing Walls

O.4. - Outlet Channel Repair/Restore, 67 ft. Wide Bottom and Bottom/Bank Armor (12” rock)

In overview, this strategy consists of capital improvements that would allow the Radial Gates structure, and its inlet and outlet channels, to tolerate average channel velocities up to 10 fps and flows up to 3,000 cfs.

These alternatives utilize reinforced concrete structures in the critical, high velocity, inlet, spillway and outlet expansion zones. And they armor the remaining inlet and outlet channel bottoms and banks with suitable rock and rip rap on geotextile. These alternatives include cofferdams, pumping to drain and bypass the channels, repair of the existing erosion damage, and construction of the new concrete structures. The total construction cost is estimated at $930,000, plus annualized maintenance costs of $10,000 per year thereafter.

Page iv

Long Term operational controls still include placing specific limits on the gate opening height in order to control the entry and discharge velocities, minimizing additional scour and erosion. But these controls double the current allowable channel velocities up to 10 fps, allowing the operator much greater flexibility to quickly drain the stored water and/or release flood flows in the desired timeframe. Since there are a wide array of headwater, tailwater and gate positions to be considered, tables of conditions and matching gate positions would be prepared for the operator’s guidance.

This combination of alternatives provides a very high degree of structural improvement to protect the radial gates from being undermined, while minimizing the maintenance of ongoing scour and erosion. However, this strategy does not completely mitigate erosion produced by either high stage flood flows through the open gates (exceeding 3,000 cfs), or by flood flows over the top of the structure (exceeding el. 590 ft.). Most importantly, this strategy provides for the continued management of the water levels in the SGA’s Waterfowl Hunting Zones, eliminates erosion due to the normal annual operation of the radial gates, and secures the structure for the long term. This strategy provides a durable, long term remedy to the observed scour and erosion.

Long Term Erosion Control Measures, Improvements and Capacity Enhancements

The recommended Long Term Measures, Improvements and Capacity Enhancements include rhe following alternatives:

1.4. - Inlet Channel Repair/Restore, Concrete Inlet Apron, Wing Walls, 16" Rock Bottom/Banks

S.3. - Add Bypass Pipes with Inlet Control and Outlet Structures, Restore Existing Spillway

E.l. - Armor Existing Outlet Expansion Zone with 28” Rock and Concrete on Upper Banks

O.4. - Outlet Channel Repair/Restore, 67 ft. Wide Bottom and Bottom/Bank Armor (12” rock)

In overview, this strategy consists of capital improvements that would allow the Radial Gates structure to tolerate water jet velocities up to 18 fps, and its inlet and outlet channels to tolerate average velocities up to 10 fps, flood flows up to 3,300 cfs, and flood stages above 590 ft. elevation. These alternatives utilize a reinforced concrete apron at the critical, high velocity, radial gates inlet, heavy rock armor of the outlet expansion bottom, and reinforced concrete armor of the spillway and expansion banks. And they armor the remaining inlet and outlet channel bottoms and banks with suitable rock and rip rap on geotextile.

These alternatives include cofferdams, pumping to drain and bypass the channels, repair of the existing erosion damage, and construction of the new concrete structures. But in addition, this strategy adds increased discharge capacity in the form of new bypass pipes, with inlet level control structures and energy dissipating outlet structures. These pipes would mimic the original ‘Seven Tubes’ structure by allowing the ‘full pool’ water levels to be discharged around the radial gates, thereby minimizing high head, high velocity discharges through the gates. The bypass pipes would also significantly increase discharge capacity, reducing flood flow erosion at the radial gates. The total construction cost is estimated at $1,098,000, plus annualized maintenance costs of $8,500 per year thereafter.

Long Term operational controls still include placing specific limits on the gate opening height in order to control the entry and discharge velocities, minimizing additional scour and erosion. But these controls double the current allowable channel velocities up to 10 fps, allowing rhe operator much greater flexibility to quickly drain the stored water and/or release flood flows in the desired timeframe. Since there are a

Page v wide array of headwater, tailwater and gate positions to be considered, tables of conditions and matching gate positions would be prepared for the operator’s guidance.

This combination of alternatives provides a significant degree of structural improvement to protect the radial gates from being undermined, while adding new discharge capacity and minimizing the maintenance of ongoing scour and erosion. However, no strategy can completely mitigate erosion produced by flood flows over the top of the structure (exceeding el. 590 ft.). Most importantly, this strategy provides for the continued management of the water levels in the SGA’s Waterfowl Hunting Zones, eliminates erosion due to the normal annual operation of the radial gates, increases discharge capacity and operational flexibility, and secures the structure for the long term. This strategy provides a durable, long term remedy to the observed scour and erosion.

Recommendations

Williams & Works recommends the implementation of either Long Term strategy outlined above in order to produce a flexible, low maintenance, durable, permanent remedy. The Emergency Measures should only be implemented if the current budget constraints are inflexible, and if the Long Term control measures will not be implemented.

Page vi

1. Introduction

The Shiawassee River State Game Area is owned and operated by the Michigan Department of Natural Resources (MDNR) Wildlife Management Division. The State Game Area (SGA) includes an earthen dam which separates the Shiawassee River flats from the Bad River. This dam is known as the Shiawassee Flats Dam, Dam ID #1819. The dam’s primary water level (stage) control structure is a pair of radial gates, mounted integral to a small bridge, which control water movement from the Shiawassee River flats to the Bad River. The radial gates structure and its inlet/outlet channels are critical components of the State Game Area seasonal flooding operations — controlling the flooding of the SGA’s Waterfowl Hunting Zones and wildlife management areas south of the Bad River.

2. Scope of Study

In March of 2008, the Dam Safety Program of the Michigan Department of Environmental Quality (MDEQ), Land and Water Management Division (LWMD), conducted an inspection of the Shiawassee Flats Dam. In his subsequent report, the inspector recommended that the MDNR:

Further investigate the erosion problem on the stream banks just downstream of the radial gates structure by December 31, 2 008. On the basis of this study, install appropriate measures by December 31, 2009.

In November of 2008, the MDMB hired Williams & Works on behalf of the MDNR to perform the recommended investigation of the observed erosion as follows:

> Study and report on the existing conditions of the outlet channel bed and banks > Conduct a topographic and bathymetric survey of the outlet channel and surroundings > Obtain appropriate soil samples and conduct soil analyses ^ Review radial gate operating conditions and hydraulics > Perform a Level 1 Scour Analysis for the channel bed > Perform hydraulic/erosion calculations for the outlet channel banks ^ Evaluate the cause(s) of the observed outlet channel scour and erosion > Identify, analyze and recommend appropriate outlet channel erosion control measures

Williams &. Works has also been contracted to design, permit, bid and oversee construction of the selected erosion control measures.

During the course of surveying the outlet channel, a large scour hole was discovered. Recognizing that scour holes usually occur at both the outlet and the inlet of a structure, Williams & Works survey staff checked for and found a mirror image scour hole at the inlet. As a result, the scope of this investigation has expanded to include inlet channel erosion control, and a thorough review of the radial gates hydraulics that has caused the observed erosion.

3. Project Location

The Shiawassee State Game Area is located within Swan Creek and St. Charles Townships, in Saginaw County, Michigan, at the confluence of the Shiawassee River, the Bad River, and several smaller tributary creeks. The Shiawassee Flats Dam Radial Gates are located in the SW quarter of Section 34 of Swan Creek Township; T1 IN, R3E, at 43° 18’ 51.4” N latitude and 84" 06’ 12.3” W longitude.

4. Stream Names, Drainage Area & Discharge Rates

As noted above, the radial gates are located at the confluence of the Shiawassee River and the Bad River.

They control flow from the Shiawassee River flats into the Bad River.

The MDEQ Land and Water Management Division (LWMD) indicates that the Shiawassee River drainage area is 622 square miles. LWMD has also provided the Shiawassee Flats Dam flood event frequency data presented in the table below. The average discharge rate of the Shiawassee River is based on a Williams & Works analysis of the nearest upstream USGS gauging station at the Fergus Road bridge, gauge #04145000, for the record period from March of 1940 through June of 2007.

Event Frequency

Discharge

(Years) (cfs) Average 457

2 3,600 5 4,900 10 5,700 25 6,800 50 7,600 100 8,300 200 8,900 500 9,800

The FEMA Flood Insurance Maps indicate that the 100 year flood stage is at elevation 594 feet (ft.), which is 2 ft. above the normal dam embankment elevation of 592 ft. The design discharge for the overall dam is the 1% (100 yr.) flood discharge, which MDEQ indicates as 8,300 cubic feet per second (cfs).

Since the normal opening of the radial gates occurs during the first weeks of December, the December ‘receiving level’ of the Bad River is important to this evaluation and any resulting design. There are not any gauging stations on the Bad River downstream of the dam, but the Bad River water level is closely linked to the Saginaw Bay and Lake Huron water level. At times the Bad River flow has been reversed by northeast wind driven seiches on Saginaw Bay. There was a USGS gauging station on Saginaw Bay located at Essexville from 1953 to 1969, but the nearest current station is on Lake Huron located at Harbor Beach.

The Harbor Beach gauge history runs from 1900 to 2009. Comparing the two gauges we found that the Essexville gauge averaged 0.11 ft. below the Harbor Beach gauge; but they tracked each other very closely.

Therefore it is safe to use rhe Harbor Beach data as representative of Saginaw Bay, and by extension representative of the lower Bad River below the Shiawassee Flats Dam. Evaluation of the monthly average December water levels finds a maximum of 581.57, a mean of 578.55 ft., and a low of 576.10 ft.

5. Radial Gates Existing Conditions

5.1. Structure Description/Observations

The radial gates structure consists of a steel beam and wooden deck bridge across an opening in the Shiawassee Flats Dam. The bridge pier, abutments, wing walls and spillway floor are all constructed of steel reinforced concrete. A pair of steel overflow-type radial gates is mounted between the concrete bridge abutments and the concrete center pier. Each steel gate is nominally 20 ft. wide by 8 ft. tall, with 10 to 12 inches of wooden planking added to the top edge.

The bridge deck elevation averages 590.75 ft. When the gates are closed, the top of wood elevation is 585.97 ft. on the north gate, and 586.29 ft. on the south gate. The spillway floor is an 18” thick steel reinforced concrete slab, with a top elevation of 577.25 ft. On the discharge side, a spillway apron extends approximately 20 ft. northwest of the bridge abutment walls, with sloped concrete armored banks rising up to elevation 580.5 ft. Steel sheeting extends down 6 ft. below the concrete slab between and below the bridge abutments at both the inlet and outlet faces; to approximately 570 ft. The elevations quoted are based on Williams & Works December 3, 2008 survey using

GPS.

Visually, the structure appears to be in good condition. The concrete center pier is in good condition, with only minor surface chips. The concrete abutments and side walls are also in good condition. There are some minor cracks and efflorescence appearing on the faces of the concrete outlet side walls near the abutment corners, above normal water level. Rust spots are beginning to appear on the painted steel beams of rhe bridge. The radial gates faces exhibit surface rust, but appear to be structurally sound. The rubber gasket seals along the bottoms of the gates are visibly damaged and in need of replacement. The side seals are in better condition. The operating gear boxes and cable lifts have been serviced recently, and are now fully functional.

5.2. Inlet Description/Observations

The inlet to the radial gates structure forms a tee intersection with the Shiawassee River channel that runs alongside the earthen dam. Therefore, the inlet channel extends only as far as the toe of the dam bank; about 30 ft. The banks are generally well vegetated, with rip rap protecting the lower 5 ft. of the banks. Little or no erosion damage is visible above the water surface.

5.3. Outlet Description/Observations

The outlet channel of the radial gates structure is approximately 230 ft. long at its centerline. At 580 ft. of elevation, the channel widens immediately from the 55 foot wide concrete lined spillway apron to the 75 ft. wide eroded earth banks. About 25 ft. further downstream, the channel widens to 85 ft. and meanders to and from 90 ft. wide before joining the Bad River. The outlet channel bends gently northward to join the Bad River at a 30" angle.

As indicated in the March 2008 Dam Safety Inspection Report, “erosion remains evident on both banks of the outlet channel”. During our site visits on August 2D, November 20th and December 3rd of 2008, bank erosion was visible as far as 100 ft. downstream of the bridge wing walls. At that location, deposits of eroded soils were visible along the northern bank. At the ends of the concrete armored banks of the spillway apron, the bank rip rap and underlying soil has been washed away and deposited about 135 ft. further downstream; leaving geotextile hanging free. About 15 ft.

southwest of the concrete armor, the earthen banks are eroded back to steep clay faces. And during low water, erosion wrapping behind the ends of the concrete armored banks is evident. See Appendix No. 1 for photos of the observed existing conditions.

During our November 20, 2008 site visit we witnessed the radial gates in action. At that time, the headwater level was at approximately 585.6 ft., the tailwater level was at approximately 580 ft., and the southwest gate was open about 2 ft. Only the southwest gate was open because the northeast gate gearbox was being serviced. The photos from that visit display the incomplete hydraulic jump, the resulting ‘roller’ waves, and the turbulence that extends far beyond the existing concrete spillway apron. See Appendices 1.6 to 1.9. It is clear that high velocity, high energy, water leaves the existing spillway apron, and presumably caused the observed channel bottom scour and bank erosion. Photos from that visit also display the large, clockwise eddy that results from operating a single gate. This backflow has eroded soil behind the end of the northeast concrete bank armor.

5.4. Bathymetry/Topography

Initially, a survey was planned of the radial gates outlet channel and the dam berm, including rhe auxiliary spillway northeast of the radial gates. The survey was performed on December 3, 2008.

Based on the field findings of a 10 ft. deep outlet scour hole, rhe inlet channel was explored for a similar inlet scour hole. Finding a 6/2 ft. deep inlet scour hole, the survey was expanded to include the inlet channel as time allowed. The data from the December 3rd survey was contoured and profiled. Appendices 2.1 and 2.2 present the Berm Centerline Plan/Profile and the Topographic Survey Plan including channel bathymetry, respectively.

On March 4, 2009 a meeting was held with the MDMB/MDNR project team to review the Draft version of this Preliminary Design Report. At that meeting it was decided to conduct a more detailed bathymetric survey over an expanded area to supplement rhe results of the December 3rd survey. The goals for the additional bathymetric survey work included:

> Inlet scour hole: A more derailed survey of its shape to confirm rhe vertical drop at the edge of the radial gates floor, and to better establish the fill volume, > Inlet channel: An expanded survey area to determine the horizontal limits of the scour hole, NE and SW of the scour hole and across to the opposite bank, > Outlet scour hole: A more detailed survey of its shape to determine the slope at the radial gates spillway edge, and to better establish the fill volume, > Outlet scour mound: A more detailed survey of its shape to determine dredging volume, J* Bad River cross sections to determine the natural river bottom shape and bottom elevation.

The supplemental survey was performed on March 24, 2009, and the resulting Topographic Survey

Plan is presented in Appendix No. 2.3.

Outlet channel static water level on December 3, 2008 was 579.6 ft. with both gates closed. On November 20'1' the water level beyond the hydraulic jump was approximately 580.6 ft. with the south gate open about 2 ft. During our site visit on August 2 1“, the static water level was just below

579.25 ft. with both gates full open. On March 24, 2009 the outlet water level was 580.74 ft.

The inlet channel of the structure forms a tee intersection with the Shiawassee River channel.

Above the water line, the northeast bank slopes at 1V:2H per the original plans. The southeast bank is steep near the bridge abutment, sloping at IV: 1H, and tapers back to 1V:3H south of the inlet channel. Immediately southeast of the bridge center pier the inlet/river channel has been scoured down to 570.8 ft. elevation; about 6.5 ft. below the spillway deck. About 70 ft. east of the center pier the scour hole tapers back up to the 576 ft. elevation.

The outlet channel of the structure is approximately 230 ft. long and bends gently northward ro join the Bad River at a 30° angle. The survey found that 50 ft. northwest of the bridge center pier the outlet channel has been scoured down to 566.5 ft. elevation; about 10.75 ft. below the spillway deck. Approximately 130 ft. northwest (downstream) of the center pier, the scour hole tapers back up to the 575 ft. elevation. At 155 ft. downstream, a mound of erosion control rip rap and sediment rises above the 579 ft. elevation. Along the south side of the sediment mound the channel floor rises to an elevation of approximately 577 ft. Along the north side of the mound, rhe channel floor elevation is about 1 foot lower, at 576 ft. Since the north half of the channel forms the ‘inside edge of the flow curve’, and the north gate had been out of service, more water flowed south of the mound, leaving slower moving water to deposit sediment along the north bank.

5.5. Soil Conditions

On November 20, 2008, three soil samples were gathered from the north bank of the outlet channel, about 25 ft. northwest of the spillway apron. See Appendix 3 for the particle size analysis report. Sample Location #1 was gathered low along the bank in the exposed, but firm channel bank soils. Soils analysis classified this sample as well-compacted ‘Silty Clay’, USCS type CL-ML, with a D50 of just 0.0032 mm, i.e. 50% of particles finer than 0.0032 mm. The second sample was gathered a few feet higher on the exposed bank. This sample was classified as Sandy Silt, type ML, with an estimated D50 between 0.025 and 0.040 mm. The third sample was gathered at the top of the bank, and included some of the surficial organic soils. This sample was classified as ‘Poorly Graded Sand with Silt’, type SP-SM, with a D50 of 0.2488 mm. The Sandy Silt and Poorly Graded Sand with Silt soils are easily eroded or scoured away by water moving at velocities above 2 ft. per second (fps). The cohesive Silty Clay soil that forms the bottom of the channel is more durable, but still generally susceptible to erosion/scour from water velocities above 4-6 fps.

5.6. FHWA Level One Scour Analysis

The Federal Highway Administrarion and MDOT routinely perform scour analyses of highway bridges. For this combination bridge and dam structure, we have prepared the Level One Scour Analysis presented in Appendix No. 4.

Regarding Data Collection, no underwater inspection reports are available. We recommend additional inspection at the onset of construction when temporary cofferdams are set in place.

Regarding Field Investigation, portions of the streambed and banks are clearly unstable and have changed compared to the original plans, and the 2008 and 2009 surveys found that:

> upstream and downstream scour holes have developed, > downstream channel banks have been eroded and have caved in (degraded), and > downstream spillway apron and abutments protection is degrading.

Also, the upstream scour hole appears to be endangering the inlet scour cutoff wall below the spillway floor near the center support pier, and the downstream scour hole may be endangering the outlet scour cutoff wall. This will be an important issue to inspect and correct during construction.

Overall, the bridge and radial gates structure itself appears to be stable, and the erosion/scour of the inlet and outlet channels can be corrected. Since the structure is stable, and it is not part of a public roadway, no Level Two Analysis is recommended at this time.

5.7. Summary of Erosion/Scour Impacts Observed

Visually, the bridge and radial gates structure appears to be in good condition with no apparent erosion impacts. However, cracks have appeared in the downstream faces of the concrete outlet wing walls near the abutment corners, and these should be monitored for further movement.

Inspection below the normal water surface has not been done, but this can be performed at the onset of construction work when temporary cofferdams are set in place.

The inlet to rhe radial gates structure exhibits little or no erosion damage above the water surface.

However, immediately southeast of the bridge center pier the channel bottom has been scoured down to 570.8 ft. elevation; about 6.5 ft. below the radial gates floor elevation. This scour hole is presumably the result of excessive radial gate inlet velocities. The record drawings show that a steel sheeting scour cutoff wall extends down to 7.5 ft. below the floor, or rhe 570 ft. elevation. We recommend that the inlet cutoff wall sheeting should be inspected during construction for perforations and/or underflow erosion near the center pier.

The banks of the outlet channel exhibit erosion as far as LOO ft. downstream of the bridge wing walls. Bank rip rap and underlying soil has been washed away, leaving geotextile hanging over the steep clay faces. During very low water, erosion damage is visible behind the ends of rhe concrete armored banks, deposits of eroded soils are evident along the northern bank, and a mound of rip rap and soil is visible in the center of rhe channel. The bathymetric survey found a large scour hole about 50 ft. northwest of the bridge center pier, with its deepest point at approximately 566.5 ft.

elevation; 10.5 ft. below the radial gates floor. This scour hole and bank erosion is the result of the turbulent energy eddies and the high velocity of the water being discharged through the incomplete hydraulic jump at the spillway. The scour and erosion is aggravated by the return circulation resulting from the unbalanced single gate discharge. The record drawings show a second scour cutoff wall below the radial gates outlet that also extends down 7.5 ft. to the 570 ft. elevation. We recommend that the outlet cutoff wall sheeting should be inspected during construction for perforations and/or underflow erosion near the center pier.

As the outlet channel bottom tapers up from the scour hole a mound of deposited rip rap and sediment is centered about 155 ft. downstream. Along the south side of the sediment mound the channel floor rises to an elevation of approximately 577 ft. Along the north side of the mound, the channel floor elevation is about 1 foot lower, at 576 ft. Since the north half of the channel forms the ‘inside edge of the flow curve’, and the north gate had been out of service until recently, more water flowed south of the mound, leaving slower moving water to deposit sediment along the north bank.

We agree with the conclusion of the MDEQ Dam Safety Inspection Report which states, “This erosion appears to be the result of high velocity releases and eddying from operation of the radial gates.” This topic is addressed further in the Radial Gates Hydraulics section of this report.

6. Radial Gates Hydraulics

The Shiawassee Flats Dam controls water movement from the Shiawassee River flats to the Bad River. The radial gates structure is the primary water level (stage) control structure for the dam. The radial gates structure controls the seasonal flooding of approximately 1,600 acres within the Shiawassee River SGA; the Waterfowl Hunting Zones and the wildlife management areas south and east of the Bad River.

6.1. Normal Operating Conditions

Typically, the gates are closed beginning September Is' of each year to flood the woodlots and waterfowl hunting zones south and east of the structure. Normal full pool stage of approximately

586.35 ft. elevation is usually reached within 2 to 3 weeks after gate closure. During the waterfowl hunting season the impoundment stage is normally maintained by the overflow rates of the radial gates. Water from the flooded areas is also pumped to other managed waterfowl hunting zones north and east of the structure. Following the end of waterfowl hunting season each year, the gates are opened to release the stored water on about December 1”. The gates are usually opened about 1 to 2 feet, and it takes about 2 to 3 weeks to discharge all of the stored water. When the upstream/downstream stages are equalized the gates are opened completely, and water flows freely from the Shiawassee River to the Bad River until the following September closing.

6.2. Flood Flow Operating Conditions

According to the 2008 Dam Safety Inspection Report, the drainage area contributing to the Shiawassee River’s flow through the dam is approximately 626 square miles. The FEMA Flood Insurance Maps indicate that the 100 year flood stage is at elevation 594 ft., which is 2 ft. above the normal embankment elevation of 592 ft. The design discharge for the overall dam is the 1% (100 yr.) flood discharge, which MDEQ indicates is 8,300 cfs.

When high river flow periods coincide with closed radial gates, water may be allowed to overflow the gates or the gates may be opened as needed to maintain “normal full pool” near rhe 586.35 ft.

elevation. Extreme flood events that cause the stage to rise above elevation 587.25 ft. will also produce flow over the auxiliary spillway; a low section of the berm northeast of the radial gates.

During these flood events the Shiawassee and Bad Rivers are typically both at high stage, so the differential head across rhe dam embankment is relatively small and rhe resulting velocities are moderate. To date, such flood flows have not resulted in any significant damage to the radial gates or the dam embankment. Therefore, flood flow conditions are not the main focus of this report, but they will be examined with respect to outlet channel erosion.

6.3. Discharge Equation Selection

The construction of the radial gates structure does not include the typical Ogee weir and discharge chute, so the standard HEC-RAS radial gate discharge equations for that configuration are not directly applicable. However, the radial gates employ curved gate faces with a flat concrete spillway, so the standard discharge equation for straight face sluice gates underestimates the discharge flow.

Reviewing the recent literature, we identified a 2006 article by M. A. Shanhrokhnia and M. Javan

(S&J) in the Journal of Irrigation and Drainage Engineering entitled Dimensionless Stage-Discharge Relationship in Radial Gates’. The article develops and evaluates two useful stage-discharge relationships for a radial gate on a flat, broad crested weir; one for free flow and one for submerged flow discharge:

Free Flow Condition: K = 0.88-z-(He/z)°*

Submerged Flow Condition: K = 1. H-z^He/z)0 ”

Terms: K - critical depth, Z - gate opening, Ht - upstream energy head

From the critical depth, the critical velocity and flow can be calculated directly. We compared the results of the HEC-RAS sluice and radial gate equations to the S&J equations, and we believe the S&J equations are a better fit to the Shiawassee gates actual hydraulics.

Free flow and submerged flow tables for a single gate with openings from 0.5 to 9.0 ft., with headwater levels of 2.0, 5.5, 9.1 and 10 ft. (flood stage) above the spillway floor, are presented in Appendix No. 5. For submerged flow, the tailwater levels used are the estimated distinguishing conditions, or submergence transition points, for the indicated headwater level and gate opening.

The appropriate HEC-RAS equations were used to estimate the initial flow needed to begin iteration of the S&J equations. For free flow, the S&J equations matched within 10% of the initial HEC-RAS equations. So, just a few manual iterations were needed to match the estimated flow (Qest) to the calculated flow (Qcalc). producing results within 0.2%. For submerged flow, the initial HEC-RAS equations varied more widely, but several manual iterations produced results within 0.2%.

6.4. Design Hydraulic Conditions Selection

In order to develop the Basis of Design, applicable hydraulic conditions must be selected. The selection process for the radial gates involved a detailed review of the radial gates’ discharge flows, velocities and energy conditions (Fronde numbers) over a wide range of headwater, gate opening and tailwater conditions. To do so we calculated and evaluated the following:

> Radial Gates Distinguishing Conditions (i.e. submergence transition points), > Radial Gates Flood Stage Submerged Discharge Conditions, > Radial Gates Normal Stage Submerged Discharge Conditions, > Radial Gates Normal Stage Free Discharge Conditions, > Inlet Channel Hydraulic Conditions, and

> Outlet Channel Hydraulic Conditions.

For brevity, rhe details of the design hydraulics selection process are provided in Appendix 5. The results of the process are presented in the Basis of Design Section 7, below.

7. Erosion/Scour Control Measures - Basis of Design

7.1. Inlet Channel Hydraulic Conditions Selected

Based upon the review of the various radial gate operating conditions presented in Appendix 5, the most erosive inlet channel conditions occur predominantly during rhe discharge of stored water through the partially open radial gates under normal free flow conditions. Based on the Outlet Channel Hydraulics evaluation, the maximum High Headwater Free Flow through the radial gates will be limited to 1,500 cfs/gate with a gate velocity of 16.15 fps; resulting in a vertically averaged Inlet Channel velocity of 8.25 fps. Also, rhe maximum Flood Stage Submerged Flow through the radial gates is estimated as 1,650 cfs/gate with a gate velocity of 10.0 fps; resulting in a vertically averaged Inlet Channel velocity of 8.25 fps. Therefore, design of the inlet channel repair, regrading and erosion control measures will be based on the following range of high head free flow and tlood stage submerged flow hydraulic conditions, and related channel design criteria:

> Vln|et = 5.0 fps for Q = 200 cfs/gate, 2 ft. of headwater, 1.2 ft. radial gate opening, to > Vlni„ = 8.25 fps for Q = 1,500 cfs/gate, 9.1 ft. of headwater, 4.65 ff. radial gate opening > V,n|et = 8.25 fps for Q - 1,650 cfs/gate, 10 ff. of headwater, 8.25 ft. radial gate opening > Inlet Channel Design Bottom Elevation = 575.25 ff. (2 ft. below rhe radial gate floor) > Inlet Channel Design Armoring Elevation > 587.25 ff. (effective flood stage) > Maintain Original Design Inlet Channel Bank Slopes of 1V:2H

In addition to the vertically averaged Inlet Channel Velocities listed above (VIr,ict), the design must consider locally higher velocities that may occur along the bottom of the inlet channel. The following Gate Velocities (VGate) represent the maximum expected inlet channel bottom velocities:

^ VGate = 10.0 fps for Q = 1,650 cfs/gate, 10…