Nova Kakhovka Dam: Erosion vs. Explosion Debates

Part 6 of "What Caused the collapse of the Nova Kakhovka Dam?  "

August 8, 2023

Limited Truth vs. Anti-Truth Cultists 

In 5 bloated posts here, I've been making the case that Ukraine and its hydropower agency UHE were the ones who destroyed the Nova Kakhovka dam two months ago now. (for reference as they come up - part 1 - part 2 - part 3 - part 4 - part 5) The story is of less interest to most people now, and that's lucky for Ukraine. Much evidence I've uncovered assembles nicely into a case I could see almost from the start - Kyiv consciously weaponized the Dnipro River against the Russians, destroying the occupied Nova Kakhovka dam with deliberate over-flooding in April and May, atop erosion worsened by Ukrainian rocket attacks, deepening this hydrologic scouring until it undermined the dam. 

I haven't proposed this a certain fact to fit my preconceived notion, but as a well-illustrated case to consider. I strongly suspect it's true, as the evidence I see keeps fitting with my starting hunch, or a version of it that grows to fit the evidence (hopefully more than the other way around). It's a familiar process by now. As always, I don't know the truth for certain. And to others, I can only make the case as best I can. 

One part of my theory is likely to become a mainstream "established fact," even though it's been loudly overruled so far. Erosion, not explosions, probably brought down the dam. Ukraine's celebrated post-2014 government, its military, and its dam operators were always clear a Russian bomb here or there was responsible. Supportive governments, Norwegian experts with seismograph readings, US spy satellites, experts cited by the New York Times, and the entire corporate media - mainly citing the Times piece - all claim to prove the same. That's the anti-Russia orthodoxy. 

And yet quite a few mainstream-thinking Ukraine boosters - especially the ones who know the evidence and the related fields - have been surprisingly unhesitant to question that orthodoxy and propose erosion was the likely cause. They have no problem holding Russia to account for alleged crimes, but this time the evidence leads them to let it go. They generally conclude that Russia is guilty, but of something more like negligence than terrorism. Still, I find it refreshing - a breath of less-stale air.

First, I noticed Ryan McBeth - an "OSINT" analyst of moderate knowledge and middling talent, with some mutable respect for reality - was willing to basically dismiss the bomb narrative, and suggest the dam fell due to erosion. Link to research, with discussion: Ryan McBeth on Twitter: "Explosives didn't bring down the Kakhovka Dam, negligence and Colin Powell's "Pottery Barn Rule" did. - Video: How Russia Destroyed the Kakhovka Dam - YouTube - The video was a starting point for my own analysis in part 1.

Regarding the seismic signals, McBeth spoke with a seismologist, Prof. Eric Dunham, and came away with the impression that if there were bombs used, "the explosion AND subsequent collapse should come in as two different kinds of waves," while we see a concentration of spikes that could just be the rumble of the dam breaking and the start of violent flooding. Dunham, McBeth, myself, and many others agree with this, and at least one of us is an expert. 

As many others have done, McBeth also pointed to the curved section of roadway that collapsed a few days before the dam, as a sign of progressing erosion (my own graphic above, and McBeth narrowed the collapse down to in between images of June 1 and June 2). He didn't seem to understand the concrete apron making this an unlikely event, but then he also missed the signs of failure in this apron making it pretty likely after all (see my part 4). He didn't know exactly how, but one way or another, he was sure some Russian negligence was to blame.

Ukraine-boosters were fairly polite in replies to this work, since McBeth was blaming Russia, after all. But they were sure he had it wrong, that the seismic readings proved a deliberate and criminal explosion. They couldn't demonstrate why, but that's what they had heard. Later, the NYT report would erase all doubt for such people.

For some reason, Eliot Higgins and the crew at Bellingcat have barely discussed the dam collapse, that I found, except indirectly. They used OSINT to identify a new dam the Russian built at Tokmak, they think to cause new flooding and prevent a Ukrainian encroachment. They put this out as indirect support to Russian dam plots, but had little to say about Kakhovka itself, aside from this side-note tweet: "The exact events leading up to the collapse of the Nova Kakhovka dam remain unclear. @nytimes reported that evidence suggested it was instigated by an inside explosion set off by Russia. Bellingcat has not been able to independently verify this claim."

Wolfsangel-clad Pro-Ukraine cultists, the same Banderites already angry at Bellingcat for supposedly exaggerating Ukraine's Nazi problem, showed up to pile on the hate for this skepticism. Why was it even needed to "independently confirm" the evident facts? Everyone knows Russians mined the dam, planned to blow it up, admitted to blowing it up, and satellites and sensors picked it up, so how dare they question NYT? If you don't blame Russia unquestioningly for the maximum crime, you're basically killing these people all over again. And so, according to these people, Bellingcat supports Russian genocide and ecocide in Ukraine. They might be working for Russia, consciously using their other work as a disguise for that core mission. 

The nerd has me blocked, but I hear Bellingcat member Aric Toler suggested an act of "mother nature" might have collapsed the dam, rather than anyone's bombs, although he says "I'm the furthest thing from an engineer." I've seen a lot further from an engineer than that. He was piled on by the cultists, I guess leading to their stretched reading of what the Bellingcat account had said (see here and here and the cartoon above). 2 strikes plus the Nazi talk and it was looking pretty obvious Bellingcat were Russian stooges. To these people. probably 98% of people on Earth would classify as Russian stooges.

Andrew Barr is not quite an example of someone agreeing with me across the aisle, but he's worth a mention. He did some good work\on the dam collapse I hadn't looked at closely enough until just now, including a look at erosion as "worth discussing" at least. He looked at the 2017 failure of the Oroville dam, noting "It's possible that something similar could happen at Kakhovka, but the drawings suggest great care was taken to account for the ground conditions." (tweet) That's the concrete apron. He wondered if the roadway collapse was caused by erosion, but seeing this, he decided probably not, even though that remains the best explanation. My own analysis suggests the apron was broken and uneven well before 2022, a condition worsened by the war and the constant, narrowed outflow, and that its worsening is just what led to the subsidence of a part of a divider and then this roadway, with its attached columns and flow guides that probably went first. A Barr commenter raised some excellent points including this, also noting "I happen to live with two civil engineering majors.  Both came to the conclusion it was some sort of scour [erosion] failure after seeing that the bridge supports were gone the morning before the dam gave way." (these were actually gone earlier, on June 1 or 2).

Barr would decide "I think internal explosion(s) remain the most likely cause," even though "nothing I've said rules out existing damage to the barrage and downstream erosion as comorbidities." (Andrew Barr on Twitter.) As far as I can tell, the kind of structural collapse he suggests at the HPP could be caused by an internal bomb or a mechanical collapse as I propose. I think he could and should come back to the issue. 

I had some agreeable interactions with Dutch researcher OSINTJOURNO - a Ukraine booster who nonetheless saw signs of erosion perhaps related to Ukrainian rocket attacks, perhaps in addition to explosives. She followed this with a rigorous analysis of all images, construction details, as well as plant employees, although the latter proved the less reliable evidence in the end. A later review thread with great visual aids said in part; "It cannot be ruled out that the Russians may have accelerated or aided the collapse of the Kakhovka dam on June 6" but "the dam was already so structurally damaged on June 2, 2023, that it was no longer salvageable at that time." As I've put it, there may have been bombs set, but if so, it comes with this coincidence that the dam was set to collapse any day regardless.

Debate with Julien DuPont

My most interesting point of agreement came in a less agreeable discussion with Julien DuPont, who claims and seems to be a French expert in the field ("Hydrology and hydraulic infrastructures happen to be my job." tweet). He responded to my last article, starting with: "Fully agree that the destruction was caused by erosion and not explosion. Fully disagree that Ukraine had anything to do with it: time series of water levels in the upstream dams show that their dams management was perfectly normal and similar to past years." 

As I started out, "Similar" doesn't cut it when the differences are what they are - skipping spring irrigation 'til mid-April, taking on heavy rains then w/net DROPS after, as Kakhovka was flooded with all of the excess, after all that damage, stuck gates & erosion ..." Nothing DuPont said really addressed that, but we repeated the exercise several times, and there was a chance for learning on both sides - along the way some new data and thinking have already re-shaped my views slightly. 

DuPont acknowledged Ukraine may have made some minor mistakes ("Neither side had any interest in destroying the dam. But neither side did its best to prevent this from happening. Very sad story." tweet), but if anyone was to blame, it must be Russia. He came with assumptions of normalcy. and the approved mainstream bias you're supposed to ignore, thought he could set me straight with some basic words, with no original thought or learning required on his own part. Then he had a hard time accepting that wasn't the case - I had questions he couldn't answer so easily, and plausible answers he didn't even try to understand. 

He dabbled in false-flag science to bolster his preconceived notion that I was totally wrong. He doubted the Ukrainians would waste valuable water by destroying a dam and its reservoir and so, if they wanted to do it, would chose bombs instead, and would have been able to do that (tweet). But if they did decide to use the river itself, he's sure they would have wasted way more water than the amounts sent, even though he agrees that did suffice to collapse the dam (tweet). He knows water, and sometimes logic.

He rightly pointed out that the small reservoirs, which I had considered too heavily, barely matter for volume; comparisons between the 2 big lakes - Kremenchuk & Kakhovka - tell most of the story more simply. He focused on Kremenchuk levels, saying over and over how they were similar to those in most years, and would come out much lower if Kakhoka had been over-flooded. But he seems to ignore the nature & timing of the small differences with other years and, more importantly with the other lake this same year, getting out-of-phase with Kakhovka in a way that really illustrates my own points (see graphic with red and blue lines below).  

DuPont: "If it was really UKR's objective to flood the KL, then they would have doubled down and released massive amounts of water. That did not happen, as the Krementchusk level remained constant until June."

Me: "Hey, they doubled down and released massive amounts of water. That's why Kakhovka was 100% full, 1m past normal max, for one month, despite 4 gates pouring 24/7. It's a crazy lot of water. Not seeing it vanish somewhere else doesn't change that." 

That was my favorite point. He fails to understand we're considering a possible shell game with water, where movements can be obscured, where outflow can be replaced to maintain the natural appearance, especially when there was heavy new flow in April to draw from. And that natural appearance matters, as DuPont knows, having cited it right off to suggest Ukrainian innocence. He could make better use of the volumes and timing of that game, if it existed. But he only wants to squint away such details until everything looks fuzzy but kind of normal, just like he lazily anticipated. 

Freeboard?

DuPont also argued that Lake Kremenchuk was filled roughly to the physical brim at certain points over the years, and that's why it's unlikely would they keep a significant "freeboard" of unfilled capacity that might have been used to forestall disaster in 2023: "the maximum level observed 4 times in 30 years HAS to be very close to the maximum allowable level. And obviously, the maximum level is NEVER to be exceeded." (tweet). He tried it out for good measure, though he didn't seem to think it was real: "But let's assume there was, let's say, 1m available freeboard that could have been filled." (right) He decided it wouldn't have mattered: "by 9th April, the Krementschusk reservoir would have been full and the UKR would have released water to keep the level constant, in exactly the same amount as they did in April-May." They would do it later, after Kakhovka had time to deal with the excess it was just dealt. That alone would have made a huge difference.

He's the expert, but I read that some freeboard is a standard safety feature of dams. Freeboard - Types of Free Board, Determination of Freeboard & Its Uses (aboutcivil.org) explains that it's usually 1.5 times the height of expected waves atop the "maximum design-pool elevation." In practice, I only know Lake Kakhovka normally kept 1m of freeboard;16.5m is the normal maximum held for years, and it was filled the last meter to 17.5. before it started pouring over the top. I don't think these figures are in dispute. Kremenchuk and the others are quite likely the same, despite what DuPont thinks, although I suppose we can't be certain of that. 

Logically, you wouldn't use this extra capacity lightly, but heavily, when the alternative is dangerous flooding downstream, like forcing Kakhovka to use ALL of its "freeboard" for a month straight. And that applies here. 

DuPont linked me to a new resource, a USDA site with reservoir levels read by passing satellites. By the USDA graph, I'd say the normal high at Lake Kremenchuk is 81.6m (about what he gets - see above). 81.8 was reached last year, and about 81.9 this year, both exceptions to the norm, bur proving there is at least about 0.3 meters that could be filled, yet never had been until the current conflict. 81.9 could be the physical limit, but we're considering freeboard, and the limited precedent says it may be 1m above the normal high. That would be 82.6m. This year they filled it to or past normal, to 81.6 or 81.9, depending on the measure (USDA chart gives the higher number, Hydroweb stations the lower one). That would leave 0.7m to the full meter of unused capacity, depending. I'm going with 0.7 and suggesting they used some of that space this year - about 30% of it. But again, this is all guesswork.

Lake vs. lake

Next, I compared the USDA graphs for the 2 lakes (G-REALM - Kremenshugskoye (usda.gov) & G-REALM - Kakhovskoye (usda.gov)) with 2022-23 overlaid so the known and possible 1m freeboard are to scale and therefore all is - 1m = 1m. This puts the lie to this claim of harmonious balance as in years past. 

The reservoirs start roughly in sync, but from mid-2022 as Ukrainian attacks on the dam commenced, a pattern emerges where Kremenchuk drains extra low as Kakhovka fills extra high, or in Dec-Feb, it holds back so that Kremenchuk was building to its normal high through February and March, as Kakhovka sank to a record low in February, and was kept well below normal range through March. You build up like Kremenchuk did by taking water in and passing on little to none of it. Normal maximum and record low, on the same river, just 1 month apart - that's a disastrous imbalance. Some of that water should have already been sent to Kakhovka and passed into the sea in an orderly manner. Instead, it was saved up all winter (this, noted skipping of spring irrigation at Kyiv HPP) as Ukraine complained of the shortfall at Kakhovka. It was kept until 20-year record rainfall was on its way, and THEN it was sent almost all at once along with that rain.

My big point in part 2 was how this heavy water load should be shared better than it was. I've learned some details since then, and now I see it was shared a bit better than I thought. I'm noticing in the USDA chart a serious drawdown at the end of March that was only faintly suggested in the Hydroweb records. This would serve to refill the stations that had apparently just raised Kakhovka to low normal. (Dnipro HPP had just begun its filling that would never stop, but at the moment, it was some belated correction of the imbalance). They likely could foresee the heavy April rains coming and made some room. 

Still, when that rain came, it pushed both lakes past normal high levels. But there was still a serious imbalance in that. Kremenchuk's level rose 1m from early April, and spent some 50 days past normal high, mostly at about 0.3 or 0.4m past. That's nothing to scoff at, but they probably retained 60-70% of their freeboard capacity at the worst point. Kakhovka rose some 2m after early April and might have risen further if that were even possible. It spent about the same 6-7 weeks past its normal high, and most of that time was spent 100% full. 

Retained likely freeboard: 60-70% compared to 0%. That's a load that could have been shared better, Wasn't I supposed to learn something new here? And from there, Kremenchuk shed some, and was about 0.1m below normal high and probably 1.1m below physical capacity by early June, as Kakhovka was wrapping up a full month AT physical capacity, and was about to collapse. 

DuPont: "Any dam can store water up do a maximum level determined by its design. A proper dam operation in case of intense rainfall consists in keeping the water level below that max level which is obviously the upper end of normal range. It should never overtop." 

Me: "Well, UHE thought Kakhovka should overtop for a month straight. Otherwise, they would NOT have kept sending ALL the excess to them. They would share the load, if they agreed with you. And usually I guess they do. But seems they had special considerations this year."

Floodgates? 

According to the data, DuPont feels the amount of water sent by UHE's dams in April and May was just fine, and only became disastrous because of the Russians. Safety standard left no room for excess upstream, so UHE had to send it. Kakhovka could have made room just fine, but the Russians messed that up by opening too few floodgates. Therefore, I gather, it's OK that UHE's dams filled Kakhovka to 100% full and MAINTAINED it like that for weeks on end. 

It's just like those traffic laws that apply when you drive your bus into a car that should have turned out of your path but never did. You saw it coming and could have turned yourself, but the other driver fell asleep at the wheel or whatever, so it's OK to drive right over that car. In fact it's the responsible thing to do, since turning can be dangerous. In the same way (?), DuPont maintains if the Russians wouldn't or couldn't open more floodgates, then it's fair to send such a torrent it cannot be managed, and to see that happening for weeks on end as you keep adding to it, until you collapse the fucking dam with erosion. Pardon my French. 

Then you can say "well, they should have opened more (and/or different) floodgates." This is what "responsible dam operators" do. Or so I gather, after some discussion with an expert. 

Now, obviously, IF the Russian-affiliated dam managers could open more and different floodgates, that would lessen the overload, while adding less to the worsening erosion. They quite obviously should have changed that situation, as early as possible, once the flood was upon them. But again, it's not clear they could do this, when the cranes that run the gates never moved for 6 months - they might have become somehow deactivated. 

And furthermore, the four gates left constantly open were already pushing the downstream areas to near flood level. Any more release, as DuPont proposed, would probably push it past flood level. 

Not because they wanted to, but because they were trying their hardest to drain the reservoir OR doing the NOTHING they could do, the Russians were already flooding their own defensive trenches, reportedly drowning one soldier. On hearing me point that out, DuPont almost giggled and suggested the Russians COULD just leave, because that flooding really should have been worsened ("If they didn't want to be drowned, nobody prevented them from leaving the area." tweet). I reminded him people live there too, and he may have dropped that point, tacitly ceding that flooding the Kherson way was not the ideal solution. Opening more floodgates would entail that, so it was never the right answer to a situation that really should not have existed. 

So even at its best, his reason to approve of this over-flooding lacks real value. Ryan McBeth, Aric Toler and the rest will probably hold similar flawed assumptions as they ignore how the situation at that dam was made unmanageable ... let's just say "for no obviously wholesome reason." As I noted early on, the Russians were left releasing far too much water AND not releasing nearly enough water all at the same time, having ended up with way too much to deal with. I'm not crazy for seriously wondering about that. 

Maybe he can come around, but it's difficult for some people. At the end he was striking a fake or uninformed pose of dismissing my whole case as "a joke" based on "pre-established conclusion that UKR had an evil plot from the start" (tweet), only imagining that I had found supporting evidence. I don't believe it when he says he fully reviewed my bloated writings before declaring global fraud, all comprehensive and proven-like: "I read everything, including your blog posts, very carefully." But he showed unawareness of my points on several occasions and, as noted, failed to really address the relevant evidence, even as he pretended to. 

K Johnson was on hand to like and maybe advise DuPont's last tweets, It was more his poser style than that of the human being I was interacting with previously. But perhaps that's just him being frustrated with me. 

So anyway, while other people swap NYT/SBU ghost stories about Russian bombs, give each other prizes for the best story, and shame anyone who doubts the stories, that's the kind of lively debate that exists among smart people discussing what ACTUALLY happened to the Nova Kakhovka dam.