Vernon County and the counties that surround Vernon County have had a love hate relationship with the earthen flood control dams that were built here in the 1950s, 60s and 70s. While they have certainly prevented flooding and damage in the decades that they were in place, the dams, known as PL566 dams (PL stands for public law), were not built for the increasingly severe weather events, are all at or beyond their 50 year life expectancy, have been expensive to maintain, and a recently completed study by the federal government shows they have likely cost more than they are worth.
Vernon County has 22 PL566 dams that were built by the federal government and then turned over to local units of government for them to maintain once they were constructed. 22 dams is about 25 percent of the 88 dams in the whole state of Wisconsin in one county. Monroe County has seven PL566 dams in the Coon Creek watershed and La Crosse County has two PL566 dams, bringing the total for the three counties to 32, or about 36 percent of the dams in Wisconsin. Nationwide there are 12,000 PL566 dams with the highest concentration in the state of Oklahoma.
The federal government agreed pay for the initial cost to build the dams, but with a requirement that 50 percent of landowners in those watersheds have conservation plans in place. That requirement was crucial to the life of the dams as it prevented excessive soil runoff that would quickly fill them with sediment, making them ineffective.
Vernon County has invested in maintenance and repairs to the dams in the 50-60 years since they were built. The Vernon County Land and Water Conservation office has staff and equipment to oversee the maintenance of their dams, including owning and operating a specialized hillside mower to keep the structures free of vegetation and trees. And the county has worked with state officials at the Department of Natural Resources (DNR) and federal officials in the Natural Resources Conservation Service (NRCS) to undertake minor and major repairs over the years to make sure they were safe.
That work ranged from minor maintenance, like fixing access roads, overflow pipe repairs, debris removal, all the way up to major structural repairs like spillway reconstruction/lining projects, to grouting projects like Jersey Valley in 2009. That project was $3.5 million project that involved pumping grout into the base of of the dam and the abutments (where the dam meets the hillside) in a series of rows, or curtains, to fill fissures and voids in the bedrock below the dam. That project was an attempt to stop leakage through the base of the dam and the hillside. If water is allowed to seep through a structure it could eventually lead to a weakening of the dam, and possibly a catastrophic failure.
Jersey Valley had been leaking through the hillside almost since the day it had been built, and the DNR ordered the county to draw down the dam following a manure runoff incident that killed most or all of the fish in the lake in 2005. That incident gave the DNR the opportunity to order the draw-down and repair.
Between the 2008 and 2018 flood, Vernon County spent over $12 million dollars to repair damage to several of the counties 22 dams. Some of those costs were carried by state and federal agencies, but a portion of it was also carried by county taxpayers that picked up the tab for the local cost share that ranged anywhere from 15 percent to 50 percent, or more, depending on the project and the source of funding.
Below is a 2009 video our of that repair project at Jersey Valley Dam with former Vernon County Conservationist Kelly Jacobs, describing the work being done there.
A decade of flooding forces a closer look
Things really started to change starting in 2007 when major flooding seemed to happen every couple of years. 2007 was considered a record breaking flood, and then in 2008 another flood surpassed 2007. Fast forward to 2018 and another flood would shatter numbers set in 2008. But depending on what standard you use or what yardstick, there has been 100 year, or 500 year, or maybe even a 1,000 year flood every couple of years from 2007 to 2019?
Those flood events caused various levels of damage to local PL566 dams. in 2007 Runge Hollow Dam suffered major damage to its spillway causing a major grout repair project and an upgrade to a concrete spillway at a cost of about $1.9 million.
The Yttri-Primmer Dam was also began serious leaks through the hillside as a result of the 2007-2008 floods, causing the DNR to order the county to cut a notch in the dam out of fear that it could fail if another major flood event happened. That notch was cut in the dam in 2008 and later refilled after a grout repair was done in the leaking hillside. Total cost of repairs was around $1.4 million.
There were numerous projects similar to those listed over the last two decades and many costing tens of thousands, to millions of dollars. The local cost share to Vernon County varied, with the majority of the cost, including technical assistance in the form of engineering, usually picked up by state an federal agencies, but not always. At the very least it is safe to say Vernon County has spent millions on maintenance and repair of their dams.
2018 changed everything
The 2018 flood shattered previous records and caused local, state and federal officials to realize that something drastic was going on and it was time to step back and do some reassessment. It made little snes to rush to fix dams that had been breached until it was determined if those repairs would work or even made sense financially.
To give an idea of how much worse the 2018 event was than previous events, the record set in Readstown in 2008 was 19.65 feet, but in 2018 the river reached 23.17 feet, shattering the old record by 3.52 feet. In Gays Mills, the record set in 2008 was 20.44 feet, in 2018 it was 22.31 feet, a difference of 1.87 feet. In Steuben, the record set in 2008 was 19.16 feet, in 2018 it was 19.84 feet, a difference of 0.68 feet.
Total rainfall amounts from Monday, August 27, 2018 through Wednesday, August 29, 2018 for Cashton 12.86 inches, Westby 12.03 inches, Hillsboro 11.16 inches, Ontario 9.99 inches, LaFarge 9.75 inches, Stoddard 9.72 inches, Wilton 8.58 inches. And what was really significant were the six hour rainfall totals that were up to 9, 10 or 11 inches in some places.
The extreme nature of that 2018 event caused five dams to “over-top” in the Coon Creek and West Fork of the Kickapoo. Meaning the water was so high that the auxiliary spillway could not keep up and the water flowed over the top of the dam itself. The failure of five PL566 dams in one place had never happened anywhere in the United States before according to engineers and local conservationists. That caused the nation office of the Natural Resources Conservation Service (NRCS), the federal agency that oversees the nations PL566 structures, to take notice and in 2019 funded a $1.6 million study of the those two watersheds.
NRCS contracted with several firms from across the country with specific expertise that could do a cost benefit analysis of the dams, both looking back over their lifespan, and looking forward to determine if replacing them would be cost effective. Those firms included M&E Consultants and sub-contractors Sundance Consulting, EA Engineering Science and Technology, Forest Econ, Inc., and Cornerstone Restoration.
The goal of the study was to look at the impact to the floodplain and model what would happen if they would repair, replace, relocate or remove the dams in those two watersheds. Personnel from the various firms and NRCS held listening sessions and public input was gathered throughout 2020 in Cashton and Coon Valley to help develop a scoping plan and an Environmental Impact Statement (EIS). The first scoping summary document was produced in 2020 and another summary report was produced in 2022 following more public input in 2021. The reports included all the feedback gathered from the public input sessions.
In 2021 public meetings were held and a report of design alternatives was delivered to the public for both watersheds. These reports contained numbers on the cost benefit analysis looking backward and forward. What did the dams save in terms of damage to infrastructure, homes, crops, loss of life, compared to what they cost.
Below are the numbers for the nine structures in the West Fork of the Kickapoo. The very top line shows a cost benefit ratio of 1.57. A cost benefit score over 1 means they were worth the cost to build and maintain them. A score below 1 means they were not. So a 1.57 score appears to be positive, making the dams in this watershed worth the cost But engineers who presented the numbers pointed out, most of that positive number was due to recreation, and most of that was due to Jersey Valley. If recreation is removed from the equation the cost benefit score drops to .13 and dropping well below a positive cost benefit number.
The study also pointed out that the areas protected by the dams was generally a long narrow area that limited the effectiveness of their protection. As the bottom of this table shows, replacement of the failed dams (vs decommissioning) in this watershed would only increase the area protected by 75 acres, 2,241 vs. 2.136.
The study also concluded that repair of existing dams was not worth examining because the original design that generally consisted of an auxiliary spillway located on the end of the dam, near the abutment, or the area where the dam ties into the hillside was faulty. The design directed water in extreme events to that area, and was the cause many dam failures like Jersey Valley, and the Mlsna Dam. The repair option was also not considered because the original dams were designed for much lower weather event (50 year flood or about 6.5 inches of rain in 24 hours) and they are not sufficient to handle the increase in extreme weather conditions.
The study gave an estimate of over $24 million to replace both the Mlsna and Jersey Valley Dam’s (built to higher standard) and about $6.7 million to decommission them.
Below is an estimate of what the dams in the West Fork saved or protected, or provided over their life in 2020 dollars. Note that the estimated recreational value of Jersey Valley over a 60 year period was $43.8 million.
The study also concluded that building all new dams would not be feasible because they would need to be designed to handle increase weather events and the design would need to be upgraded to one that is suited to the karst geology they would sit on that is fragile, and unstable. That would mean installing concrete “cut-off” walls at the base of the dam that would go well below the base of the dam to prevent seepage and moving the spillway to the center of the structure. Engineers said that design is incredibly expense to undertake would drive up costs well beyond any positive cost benefit ratio.
The study also examined what kind of impact alternative upland land use practices could have on slowing water, or increasing water infiltration. The slide below shows an estimate done in the Mlsna watershed, and shows the impact of such changes would be minimal, lowering the curve number from 71 to 68 at best. A curve of 38.5 would need to be achieved to match the level of protection that dam provided.
But another piece of information showed a number of small dams could provide some of the protection that one big dam did in the past. 11 small dams were estimated to reduce peak flow by 50 percent compared to 75 percent reduction the Mlsna Dam provided.
Coon Creek Watershed
The results of the study for the Coon Creek yielded similar numbers and a similar conclusion. The original 1958 cost benefit analysis shown below for the 14 structures in the Coon Creek showed a cost benefit score of 1.2, and those 14 structures controlled 19 percent of the overall watershed. That is 338 of the 1,303 acres.
And like the West Fork analysis, the Coon Creek numbers showed protection of recreation was one of the dominant benefits, but without a major recreational structure like Jersey Valley in this watershed to boost the numbers, the cost benefit number came in under in under 1 at .92. That means the cost of protection from those 14 structures in the Coon Creek over a 60 year period was more than the value of what they protected over that same time. Cost benefit for the 14 structures in the Coon Creek, if you remove the recreation variable, drops the number to .41
Similar to the Mlsna Watershed, engineers examined upland land use changes and how that would impact the Luckasson Dam area. Similar to the results in the Mlsna Watershed, land use changes have little impact on peak flows in extreme events, but small dams show more promise.
As with the West Fork analysis, engineers concluded repairing the existing dams or building the dams to a new standard and design would not be cost effective. The recommendation is to remove 14 structures over time. The cost to decommission is still significant, about $2.8 million, but NRCS officials have indicated they will cover 100 percent of those costs in most cases.
The decommissioning plan and public input sessions
That brings us to the recent release of the plan or decommissioning of the dams released by NRCS. The 1,046 page Watershed Plan-Programmatic Environmental Impact Statement for the West Fork and the 1,029 page document for the Coon Creek lays out exactly how each structure will be removed, including costs estimates for both watersheds.
- NRCS will hold a public meeting on the DRAFT Plan-PEIS for the Coon Creek Watershed at the Coon Valley American Legion Hall, 12:00-2:30 pm on Thursday, January 18, 2024.
- And NRCS will also hold a public meeting on the DRAFT Plan-PEIS for the West Fork Kickapoo Watershed at the Cashton Community Hall, 4:00-6:30 pm on Thursday, January 18, 2024.
You can also submit your comments about the plan online here.
With over 2,000 pages NRCS engineers have included an incredible amount of detail in their plan to decommission all of the dams in both watersheds moving forward. Suggested timelines show many of the structures are estimated to have another 5-12 years of useful operation left, but several dams are already breached and to get all 23 dams decommissioned in both watersheds in 5-10 years will be a monumental and expensive undertaking.
What does decommissioning mean exactly? The documents define decommissioning as “This action includes stabilizing failed dam sites and decommissioning the dams that have not yet
failed by excavating a notch that would be able to pass without impoundment the 100-year frequency, 24-hour storm event.”
The good news is, there appears to be a commitment by NRCS to cover the majority of costs, and they appear to be willing to rebuild Jersey Valley dam, with a recreational pool, and with a much lower local cost share than was first proposed. First, here is an estimate included in the Coon Creek document that shows what the local cost share is estimated to be to decommission all 14 of those structures. $88,700 , or 2 percent of the estimated $4.3 million.