05/30/2026
John Bulmer
May 26, 2026
Analysis of the Latest Surveys of Martin Dunham Reservoir Dam:
A Closer Look at the Data Behind a Decision That Can't Be Undone
The New York State Office of Parks, Recreation and Historic Preservation released 257 pages of inspection and engineering analysis on April 10, 2026. That document is the public record on which any future decision about the Martin Dunham Dam and the 98-acre reservoir at Grafton Lakes State Park will rest.
I read all 257 pages.
What I found is not a completed case for any particular outcome.
What I found is an incomplete analysis on a public asset that has served this community for 112 years, one that leaves unanswered the basic questions NYS Parks owes the public before an irreversible choice is made. Among the alternatives under evaluation is wetland restoration, which would permanently eliminate the reservoir. If that is the direction this process is heading, the public deserves to understand what it actually involves. These pages make that clear, and Parks has not said so plainly.
Three things need to be said before this goes any further.
Any Removal Decision Cannot Be Undone
If the embankment comes down, the reservoir is gone. The 98 acres of water. The fishing, the paddling, the shoreline that generations of families have used. The flood storage that sits upstream of the Quacken Kill communities. All of it, permanently gone. There is no version of this where New York State changes its mind in five years and puts it back.
That irreversibility is not an argument against ever making this decision. It is an argument for making it right. It is an argument for doing the work first. And that work has not been done.
The Dam Is Not Failing. It Has a Compliance Problem.
The engineering documents describe a structure with regulatory deficiencies. A missing Inspection and Maintenance Plan. Inoperable low-level outlet gate valves. Minor voids between concrete sections. An open joint in the spillway.
Those are real problems that must be fixed. But they are not the same as a failing structure, and these 257 pages do not describe a failing structure. The boring logs show an intact embankment. The stability analyses show a structure meeting engineering thresholds.
Parks has also attached a "High Hazard" classification to this dam, and that label is doing a lot of work in this conversation. It needs to be understood for what it actually means: it refers to the downstream consequences if the dam were to fail, the population and infrastructure in the flood path. It says nothing about the likelihood of failure. A structurally sound dam above a populated valley carries the same classification. High Hazard is a geography designation. It is not an engineering verdict on this structure.
What these documents actually describe is a dam that needs rehabilitation, not removal. Every deficiency identified is correctable. Rehabilitation is not delay. It is the engineering response these findings call for.
Three Things Are Missing From the Public Record. All Three Are Required.
This is where the process fails the public. The community is potentially being asked to absorb a permanent outcome based on an analysis missing three things that need to be addressed before an irreversible choice is on the table.
There is no rehabilitation cost estimate.
The documents include removal cost figures. Hydraulic dredging at $35 per cubic yard. $250,000 mobilization. $120 per ton for off-site disposal. Those numbers exist in the record.
A rehabilitation cost estimate does not. Not alongside the removal figures. Not in a form that allows the public to make a side-by-side comparison.
That is not a minor gap. It is the most important number in this entire process. If rehabilitation costs more than removal, the public deserves to know that and understand it. If it costs less, the public deserves to know that too. Presenting one side of a cost comparison and not the other is not analysis. It is not a sufficient basis for a permanent decision on a public asset.
Parks has also not told the public how many years full rehabilitation extends the viable life of this structure. A rehabilitation that buys ten years looks completely different from one that buys fifty. Without it, the cost case for any removal alternative is half a calculation.
There is no downstream flood analysis.
The Dunham Reservoir holds 98 acres of water at the bottom of a 9,500-acre watershed. When it rains hard in Grafton, that reservoir is attenuating runoff from nearly 15 square miles before it reaches the communities downstream. According to NOAA's National Climate Assessment, heavy precipitation events in the Northeast have already increased 55 percent in recent decades, the largest increase of any region in the country, and are projected to increase by at least 40 percent more by end of century.
Federal research by the U.S. Geological Survey found that removing flood-retention reservoirs increases downstream peak discharge by up to 38 percent depending on site conditions. The range varies by site, which is exactly the point. The only way to know what happens to the Quacken Kill during a major storm event without this reservoir is to model it. That model does not exist in the public record.
No analysis has been published telling the downstream communities what permanent elimination of nearly 100 acres of upstream water storage means for their flood risk. This is a gap the public has every right to demand be filled before any final decision is made.
The geotechnical investigation has not been documented for the public.
The structural assessment in these 257 pages rests on subsurface boring data. That data is the foundation of everything, the embankment condition findings, the stability analyses, the engineering conclusions that underpin this entire process. The public record does not clearly establish when that investigation was conducted, how many borings were advanced, where they were located, or whether the investigation covers the full length of the embankment.
For a permanent removal decision, that is not acceptable. The public is entitled to know exactly what subsurface investigation was done, when, and whether its scope is adequate to support the conclusions being drawn from it.
What the Wetland Alternative Actually Involves
Among the alternatives under consideration is wetland restoration, permanently draining the reservoir and converting the site. The documents describe what that construction actually requires, and it has not been stated plainly in public.
Picture what this looks like on the ground.
The only viable equipment access to this site is the Double Bit Path to the dam crest. One road in. No staging area. The existing boat dock cannot support construction equipment. Every piece of heavy machinery, every truck hauling excavated material, every piece of dredging equipment, it all moves through that single path. Standard dam removal construction involves sequential phases: site access and mobilization, reservoir drawdown, streamflow diversion, demolition, sediment removal and disposal, and site restoration. At this site, every one of those phases runs through one road over what industry guidance and documented project experience suggests would be a construction period of six months or more of active in-river work, before site restoration even begins.
While that construction is underway, the 98-acre reservoir that families fish and paddle today is a work zone. The Quacken Kill below it is a construction corridor.
The sediment that would be disturbed has been tested across five locations. The results are unambiguous: no pesticides, no PCBs, no dioxins, metals below NYSDEC cleanup thresholds in every sample. That sediment is stable and clean where it sits.
Understanding why disturbing it is still a serious problem requires a simple distinction.
Clean sediment locked behind an intact embankment is not an environmental hazard. It is a stable condition that has existed for over a century. The problem is not chemical. It is physical. When you dredge 52 to 55 acre-feet of fine silt through one access road, you suspend that material into the water column. That suspended sediment, called turbidity, flows downstream into the Quacken Kill. It smothers fish spawning beds, clogs gills, blocks sunlight from aquatic vegetation, and scours stream banks.
USGS documentation of comparable dam removal projects found that even following best practices, visitors should expect turbid, cloudy, dark water through the project area and downstream during active construction, and that the finest silt particles cannot be fully captured by sediment traps regardless of how carefully the work is managed. None of that disruption exists today. The act of removal is what creates it.
The analogy is asbestos in an old building. Intact and undisturbed, it is not an active hazard. The moment you start demolition, it becomes one. Easier to remove is not the same as harmless to remove. And at this site the disturbance happens in a constrained waterway with one road in and no staging area, which makes the exposure longer and the impacts harder to control.
Slope stabilization following removal must also still meet NYSDEC dam safety engineering standards even after the dam no longer functions as a dam. The engineering requirements survive the structure. Contractors pricing a job with one road in, no staging, and constrained access documented throughout these reports will not hold to a desktop estimate. That is not speculation. That is how construction pricing works at sites like this one.
The public has not been given a full accounting of what wetland restoration at this site would actually look like on the ground, how long it would take, what it would cost, or what the downstream and in-stream impacts of that construction would be during the process. Those answers should be in the public record before this moves forward.
This is not a demand that Parks reach a predetermined conclusion. This is a demand that Parks finish the work before any irreversible option is taken off the table.
Publish the rehabilitation cost estimate alongside the removal figures, with the service life projection that makes the comparison meaningful. Make the Emergency Action Plan public without needing a FOIL request.
Commission and publish a hydrologic analysis of downstream flood risk under all alternatives before any final decision is made.
Document and publish the full scope of the geotechnical investigation underlying this analysis.
These are not unreasonable requests. They are the minimum conditions for an informed decision on the permanent elimination of a 112-year-old public asset. Until they exist in the public record, this process should not move forward.
The reservoir can always be removed later, if the evidence supports it. It cannot be restored once it is gone. The public is owed the complete picture before that line is crossed.
This is not an engineering analysis. It is a close reading of publicly released government documents and publicly available federal research. Residents are encouraged to review the source documents and draw their own conclusions.
Link to April's report is in the comments.
With Save Dunham See less
