By allowing ads to appear on this site, you support the local businesses who, in turn, support great journalism.
Three counties to collaborate on watersheds study around breached dams
Nation's First Watershed Project
John Bollinger (former SCS Farm Planner; Hugh Hammond Bennett (former chief); Marv Schweers (former Wisconsin State Conservationist); and Herb Flueck (former Minnesota State Conservationist) standing beside a roadside marker commemorating the establishment of the Nation’s First Watershed Project in Coon Valley.

LACROSSE, VERNON AND MONROE COUNTIES - The legacy of the Coon Creek Watershed Project, undertaken in the 1930s to correct environmental catastrophe caused by unsustainable land use practices, continues to ring bells across the nation 85 years later. That iconic initiative, undertaken by the Soil Conservation Service (SCS), was the first erosion control demonstration project in the nation.

Environmental catastrophe
Camp SCS, Higgins Estate, Trout Run Watershed, Melrose, Monroe County. Severe gully erosion in Union silt loam, underlain with fine sand where cropland, pastureland had been in the prior five years. The gully is about 40 feet deep. Photo from 1937

Now in 2020, USDA-Natural Resource Conservation Service (NRCS), successor to SCS, will undertake a new watershed study in the Coon Creek and West Fork Kickapoo watersheds. This new study will help determine the future of the flood control dams that breached in late August 2018. The study will also support county conservation professionals in leading an updated approach to land use, erosion control and flood prevention in the context of increasingly large and intense rainfall events.

“I think everyone needs to recognize where this is all going,” Monroe County Conservationist Bob Micheel said. “What happened in 2018 might just be the tip of the ice berg as far as what kind of conditions we need to plan and design for. Right now, we’re just trying to keep up with the symptoms and we can’t keep up, never mind getting out in front of it.”

NRCS Conservation Engineer for the State of Wisconsin, Steve Becker, will share his agency’s dam failure report, and plan for the study of the two watersheds with concerned citizens at two meetings on Thursday, Jan. 30. Those two meetings will take place at the Coon Valley American Legion at 3 p.m., and at the Cashton Community Hall at 7 p.m. 

Also on the agenda will be a presentation by Monroe County Conservationist Bob Micheel about the status of the flood control dams that were breached, and updates from WDNR and UW-Madison. Vernon County Conservationist Ben Wojahn and PL-566 Dams Manager Mark Erickson will also be present to discuss the breached dams in their county and their plans to participate in the watershed study.

“The study that is planned will take into account hydrology,” Micheel said. “One of the questions to be answered by the study that will help us chart our future course, is whether we will calibrate to current trends and data or continue to use data that is 30 years old.”

A copy of the full dam failure report can be found on the website of the Monroe County Land Conservation Department at

The report examines rainfall reports, chronology of the dam failures, the construction specifications, hydrology and geology of the dams, a detailed description for each failed dam of the manner of the failure, a summary of possible causes for the failure, and recommendations about how to explore what the future of the structures will be.
Breached Dams
Map of PL-566 structures in the location of the breached dams with the highest rainfall totals during the August 28, 2018 rain event. The five breached dams are shown on the map. The upland areas draining to the sites are in the Oneota Dolomite (red). It has been estimated that 86 of the 89 PL-566 structures in Wisconsin are in the Jordan, St. Lawrence or Tunnel City formations (brown areas).

Dam Failure Report

NRCS released a ‘Dam Failure Report’ on June 26, 2019. The 160-page report started with a ‘General Description of Problem or Deficiency.’

The five dams which were the subject of the report included Jersey Valley Dam (WFK-1) and Mlsna Dam (WFK Mlsna) in the West Fork Kickapoo (WFK), and Luckasson (CC-21), Bilhovde (CC-23), and Korn (CC-29) in Coon Creek (CC).

The description of the failure reads:

“There was a complete failure of the five dams listed in this report. Late on August 27 a storm system entered southwest Wisconsin that included Monroe and Vernon Counties. Two PL-566 watersheds located in these counties, West Fork Kickapoo and Coon Creek, had structures impacted by the storms. Intense rainfall began to fall in this area around 9 p.m. and continued until approximately 4 a.m. August 28, an approximate six to seven-hour event. The projected recurrence intervals (statistical likelihood of the rain event being repeated – expressed in years), based on NEXRAD (Next Generation Weather Radar) system, comprised of 160 sites throughout the United States and select overseas locations), ranged from a low of about 300 years for WFK Mlsna to 650 years for CC-23.
Rainfall return frequency
Return frequency quanti-fies how frequently a rain event of the magnitude that breached the dams can be expected to occur – for example, once every 300 years or 500 years.

These rainfall amounts exceeded the principal spillway (spillway that allows water to drain out of the impoundment behind the dam on an ongoing basis) design storms (storm events that drove the engineering of the dams) reported on the as-built drawings. All three Coon Creek structures were designed with a principal spillway point rainfall of 4.6 inches. This corresponds closest to the interpreted value on the 100-Year 6-Hour Rainfall chart published in Technical Paper 40 (May 1961). WFK-1 was designed with a principal spillway point rainfall of 3.9 inches, which corresponds closest to the 50-Year, 6-Hour Rainfall chart. f

Freeboard of dams
The term ‘freeboard’ de-scribes the vertical dis-tance between the top of the dam and the full supply level on the reser-voir.

The recorded rainfall [of August 27-28, 2018] also exceeded the freeboard storms (freeboard is the vertical distance between the top of the dam and the full supply level on the reservoir) of the Coon Creek sites. The WFK Mlsna as-built drawing does not have a hydraulic sheet so there was no basis for comparison at this site. The top of dam elevation at WKF-1 was set according to a freeboard hydrograph applying a six-hour distribution with a point rainfall of 10.9 inches. The top of the dam is the level of ‘watertightness’ of the structure and may be the top of a parapet that is watertight throughout its length. The estimated six-hour rainfall for WFK-1 was 7.3 inches. This exceeded the auxiliary spillway (a spillway designed to allow water to pass downstream when the impoundment behind the dam is full to overflowing) design rainfall of 6.2 inches.”

Four of the five dams overtopped during this event. The one exception is WFK-1, which saw significant auxiliary spillway flow only. The total depth of the auxiliary spillway channel was 8.5 feet and the estimated maximum flow depth during the event was four feet. There was auxiliary spillway flow at the other four dams as well. The breach of WFK-1 and WFK Mlsna occurred in the auxiliary spillway area. CC-21, CC-23 and CC-29 breached on the end opposite the auxiliary spillway.”

“Auxiliary spillway flow resulted from this event at ten other dams in these two watersheds. Of those, five overtopped. None of these ten dams breached and damage ranged from minor to significant erosion in the groins and/or auxiliary spillway. None of Wisconsin’s (89) PL-566 structures had ever overtopped prior to this event, except for the WFK Pilot Klinkner structure that overtopped several times due to a design error.”

“According to witness reports, the breaches occurred in the early morning of August 28, perhaps between 2 and 3:30 a.m. Once the vegetative cover failed, each breach progressed very rapidly leading to a sudden release of floodwater at high head (when impoundment behind dam is full). The breach process was not only rapid, but complete. Erosion of the breach channel at each site extended into the foundation bedrock and the resulting breach channel was lower than the upstream sediment pool.”

“These sudden failures led to a large breach wave (surge of water draining out of the impoundment behind the dam) at each site with sufficient velocity and energy to cause significant damage and flooding downstream. Physical evidence of the size of breach waves can be found at the sites and in the breach inundation areas. On several sites the breach wave was high enough to overtop the lower half of the auxiliary spillway dike. Boulders were carried hundreds of feet downstream. Debris fields consisting largely of sand and rock covered wide areas of the valley floor and for thousands of feet downstream. Downstream roads were flooded with damage to culverts and bridges. Over 2,000 feet downstream of CC-23, an unoccupied ranch-style home was moved off its foundation. Fortunately, no fatalities were reported as a result of the breaches.”

Event chronology

The rainfall event that breached the five dams in this report occurred overnight during the evening of Monday, August 27 through early morning Tuesday, August 28, 2018. Mike Dreischmeier, Richland Center NRCS engineer, surveyed the damaged sites in Vernon County (Jersey Valley and Mlsna) during the day August 28th. The Mlsna flood pool was completely drained and was below the principal spillway inlet. The Jersey Valley dam had breached and was draining below the level of the principal spillway. Dan Gunderson, Sparta NRCS Civil Engineering Technician and Bob Micheel, Monroe County Conservationist, visited the three Coon Creek breached dams (21, 23, 29) on August 28th. All three of the Coon Creek dams had breached and drained below the principal spillway inlet elevation.

Additional rain fell during the afternoon of August 28th. This rain put additional flow through the breached embankments, slightly widening the breach openings.

A storm on Labor Day hit one week later – Monday, September 3, 2018. Again, more storm runoff widened the existing breach openings even more.


The recommendations from the committee formed to compile the dam failure report are as follows:

1. Conduct a Planning Study to develop and evaluate alternatives for each dam and the entire watershed. This study may include:

a. Complete an assessment of current resource concerns, future flood control benefits and costs to aid sponsors (counties) in evaluating what course of action best meets their needs.

b. Decommissioning by removal of the dam, stabilizing the site and completing stream restoration.

c. A redesign or relocation of the dam and all its components to current standards and specifications. Measures to effectively treat the foundation and abutments will be a necessary component.

2. Design considerations for dams that will be repaired or replaced:

a. Investigation:

i. Complete additional geologic investigation of the abutments (where dam joins the hillside) to determine direction and extent of sandstone formations and jointing.

ii. Coon Creek 41 (Dahlen Dam/Vernon County) had a similar breach without over-topping. Complete an inspection and assessment of Coon Creek 41. Review the failure report and repair design. Use the lessons learned from this failure and repair to guide the repair design.

b. Seepage Control:

i. Develop a pressure grouting plan or a slurry trench plan to cutoff upstream to downstream seepage flow (movement of ground water) and to prevent the buildup of hydrostatic pressure at the end of the dam.

ii. Blanket valley walls up to the top of dam elevation with compacted earthfill. Geologic investigation and seepage analysis are required to determine upstream extent of blankets.

iii. Construct the downstream groins with a clay liner to increase the head loss of abutment seepage and redirect ground water discharge downstream of the dam.

iv. Construct drains in the downstream groins to provide a stable outlet for seepage through the abutments.

c. Auxiliary Spillways

i. Design the dams without a vegetated auxiliary spillway.

ii. Provide a structural auxiliary spillway (using concrete or other impervious materials) to replace the vegetated auxiliary spillway.

iii. Design the dams with a ramped spillway located away from the abutments graded all the way to the valley floor.

d. Protect the downstream groins by raising the ends of the dams to provide overtopping sheet flow across the entire width of the dam excluding the groin areas.

3. For all watershed dams located in this geologic formation:

a. Use Geophysics technology to assess potential for failure of the abutments.

b. Prioritize sites for the geophysics analysis by using slope mapping in ArcMap to identify sites that have significant steep hillslopes near the dam that may indicate the pool areas are shallow to bedrock or have bedrock outcrops that could increase the risk of seepage.

c. Inspect the downstream groins for signs of erosion or material weakness and consider implementing measures listed below that protect the groins from over-topping or erosion from sidehill runoff.

i. Review the vegetation of the groins and assure a good stand of grass exists on the embankment and the abutment side of each groin. Where timber encroaches on the abutment side of the groin, clear and/or grub trees to provide better growing conditions for grass on the abutment side of the groin.

ii. Raise the ends of the dams to prevent concentrated overtopping flow down the groins.

iii. Build up the groins with earthfill to keep flow off the soils that are shallow to bedrock on the wooded slopes.

d. Review Emergency Action Plans to make sure contact information and actions planned are up to date. Review the protocols with appropriate personnel.