Did Ukraine Break the Dnipro River and the Nova Kakhovka Dam?

June 21, 2023

(rough, incomplete, updates likely)

(a lot of minor edits June 22, adds 6/24, 6/25, 7/10)

Introduction

My last post - What Caused the Collapse of the Nova Kakhovka Dam? - noted how, in its final weeks, the dam was dealing with far more water than it should have been. I wondered if the Ukrainians had deliberately tried to destroy the dam by flooding it with insane output from the dams they controlled upstream. Having looked into this, I'm now far more inclined to suspect over-release and erosion were key factors in the disaster, and that it was the result of such a plot, and not from anyone's attack with a bomb or missile. 

This detailed infographic shows much of what I'll explain below with some data put in some context. Six hydroelectric dams control the flow of the River Dnieper (Dnipro), and with this control could even weaponize it, as Russia has been accused of doing and as it seems Ukraine just did. 

Put another way but more basic, on a slide David Helms used in his Twitter thread:

My analysis will start with Lake Kakhovka and then go downstream to prove the overload at the south end. Then we'll move on to where all that water came from - the Ukrainian-operated dams upstream. Just which dams held, accepted, or passed on what amounts of water is not all clear, partly due to widespread missing entries at the most crucial times. But the available data generally suggests they all passed on any new excess they took in April-May, along with a bit off the top (generally about 0.2m) from what they had in late March, all of which accumulated at Lake Kakhovka, filling it up even as it flooded the downstream trying to prevent that. 

The sequence of events I propose COULD hypothetically result from natural mistakes in flood management culminating in tragedy. And it COULD just get passed off that way if the issue is ever acknowledged. But I doubt that this all came about by accident or, as the Ukrainians claim with the UTMOST CLEVERNESS, did it result from the Russians' own mindless and self-destructive brutality.

Reference:

Dams in the "Dnipro Cascade" system (Wikipedia), all hydroelectric power plants (HPP) listed north to south: Kyiv HPP - Kaniv HPP - Kremenchuk HPP - Kamianske HPP - Dnipro HPP (at Zaporizhzhia) - Nova Kakhovka HPP. The last was of course operated by the Russians, and the others are all Ukrainian-operated. As I gather, the two big lakes at Kremenchuk and Nova Kakhovka each - usually - contain close to half the water in the system, while the other three just carry a minor portion divided. 

Data source: Hydroweb (theia-land.fr) - the same one everyone's been citing as the definitive measure of water levels at Lake Kakhovka. There are Hydroweb virtual stations on the Kremenchuk & Kakhovka reservoirs, some on the river upstream from them, and others at or near the mouths of rivers opening into the reservoirs, which should read at about the same level. You don't get a straightforward reading of how much water comes in or goes out, other than by comparing levels; one gallon less than before can mean nothing came in and a gallon was sent out, or it could be that 100 gallons came in and 101 were passed on. And the times in between readings are broad - at best every ten days, most of them every 28 days, and in most cases under study, again, whole months worth of data is missing at the key times, complicating my task here.

NAFO "fella" and professional meteorologist David Helms authored a Twitter thread refuting Mikael Valtersson (or "Vatnikersson") with some useful info and misleading interpretation. He did this, as all NAFO "fellas" do everything, to maximize the return on his donations to a Georgian mercenary death squad (NAFO dues go - when they go where they're SUPPOSED TO - to the Georgian Legion. Last I heard, they were sworn to illegally execute every Russian POW they take). I'll cite "Helms" below and reply occasionally. If it's not otherwise cited, it's from this thread.

The River's Heavy Load

One bit I learned from David Helm's thread: heavy rains in April. "Past two weeks accumulated precipitation (11-23 April) indicates 20-30 mm of rainfall" There were indeed widespread flood warnings and some flooding in parts of Ukraine in those days. It was expected to peak on April 22 and perhaps did. Some reports are below, and note there were also heavy rains at least in late May, leading to May 28 flooding in Dnipro city. All of this would flow into the Dnieper River, the last fairly close to the Kahkovka reservoir in the last, fateful days before it drained away.

But here's another bit few know about, and that I don't know enough about. April 10 reports: "In Belarus near Gomel, near the border with Ukraine, due to heavy rains and rivers overflowing their banks, a dam broke through: several settlements are under water." (Dmitry Solenko on Twitter - photo source) "A dam has "failed" near Zhlobin, Gomel Region, Belarus. The impacts of the flood will eventually reach the Dnipro River entering Ukraine near Chornobyl moving southward towards Kyiv and eventually through to the Black Sea." (OSINT Uri on Twitter) Eventually - if it even existed - this flood would reach the Black Sea, but there were some twists on the way, and it would carry houses, pets and debris in the end. What a convenient turn for anyone's potential flooding plan. 

Follow-up discussion notes a lack of sources, info, or dams near Zhlobin. Žlobin on Google Maps - some 200km north of the Kyiv HPP reservoir. I only found an April 7 report with likely a different episode of flooding in some other town(s) somewhere on the Sozh River - a Dnieper tributary - where a poorly-managed "dam made of sand" had collapsed. "By April 7, the water in Sozhe had risen above the dangerous level by 108 centimeters." Only minor, shoe-deep flooding is shown in some streets and yards. 

The cause and effect is also unclear in this case, or these cases - did a sizeable reservoir drain into the river, or did the river just swell with rain and actually drain some of it into these areas, lessening the burden bound for Ukraine? Whatever sized starter flood "impact" ... it would over the following days come down the Dnieper across the Ukrainian border near Chernihiv. (FWIW It doesn't pass the Chernobyl plant - that's a different river merging with the Dnieper further south.) This and heavy rainfall on Ukraine would combine to seriously raise levels all along the river. 

It's probably wise and normal to spread the load evenly in such a case, to keep any one dam from being maxxed out or any one area from flooding. But as we'll see, the rise was artificially concentrated southward to bear on the Russian-occupied Nova Kakhovka dam. Probably with full and criminal intent, but possibly by some terrible series of mistakes, it seems the Ukrainians broke the river and the dam just had to follow suit. 

Nova Kakhovka Reservoir/Lake Kakhovka

As I gather, this expansive but not terribly deep lake normally holds close to half the water in the Dnipro reservoir system. At the end, I'd say it held ... way over half, anyway. 

Lake Kakhovka page on Hydroweb: water level normal range: 15.5-16.5m year-round, max. allowed 17m once, briefly. In 2023, levels fell to a bizarre 14.06m on Feb. 10  - then at an alarming speed, levels rose to 14.84m March 31, 15.5 by 4/11, and passing the normal range to 16.8m by April 25, then to a dreaded 17.5m by May 8. Here it's 100% full and stays about the same until the end, one month later. A slightly lower level of 17.36m is maintained, aside from a spike 17.54 on May 20, and it's even lowered to 17.26 at the end, on June 4.  (The sharp drop after the dam collapsed is considered below. It got weird later)

That's increased output at the end, but mainly - by dipping BELOW the max so we can see some small changes - this shows a rough parity between input and output for that whole month. And as I'll show, output was very high. That means input was very high, and we really should ask WHY, considering how disastrous that was getting. This situation didn't develop suddenly; people - notably at Dnipro HPP - had time to realize the danger and to keep adding to it anyway. 

KM0320: R. Kins-ka: this station is about 8km from the reservoir (Coord: 35.384, 47.7004 (lon, lat) ), not the visibly swollen stretch but still levels should be similar to the reservoir. Normal range: 16-19m. It rises to 16.6m between March 9 and April 5, then a missing entry in early May, and then 17.36 on May 29 = ^0.76m 4/5-5/29. That's still within norms here, but the trend clearly reflects the swelling reservoir, which was never a secret anyway.  

"Missed Opportunities" to Drain the Lake Kakhovka

So that was a sorely overfilled reservoir, as some astute people noticed at the time. David Helms noticed it. As he would later complain in his post-collapse thread, the Russians could have released more of that water, lessening the danger, but they blew every chance in a criminally negligent manner. "RF missed opportunities to divert water from Kakhovka Reservoir through: Kherson & Zaporizhzhia Irrigation Canals - K-HPP Lock - K-HPP Turbines - K-HPP Sluice Gates." Let's consider each of these in chronological order.

"Russian occupiers could have discharged water through the K-HPP lock, especially after the RF 11 NOV 2022 retreat, but they didn't." Why they could do this "especially" after November 11 is unclear. They built a new bridge over the lock in or by October because the Ukrainians blew up the old bridge. Made in haste, this was supported with fill that blocked the lock. (see photo: File:LockCanalBlocked.jpeg - A Closer Look On Syria (shoutwiki.com)) I didn't get all the details, but the Russians could NOT use it for emergency release afterwards, thanks to Ukrainian aggression and the questionable "Russian" response that blocked one drain. Having seen that blocked drain, someone set to blocking others.

"Russian occupiers could have discharged water through Kakhovsky Main Canal," Helms complained. But they didn't do that, he thinks, because they wanted the maximum possible volume of water there. But they didn't just refuse to pump the water or switch the station off. No, "in November 2022, the Russian occupiers bombed and flooded the KMC main pumping station creating "a disaster". Dec. 1: Base pump station within Kakhovka Main Canal flooded due to enemy shelling – Khlan (ukrinform.net) 

Obviously, they would blame Ukraine. Nov 12: after "numerous attacks by the Armed Forces of Ukraine" the hydroelectric plant was disabled and "other large facilities were also damaged, for example, the North Crimean Canal and the Kakhovka Main Canal." https://t.me/readovkanews/46721 Helms was able to dismiss these claims somehow. He didn't mention if the Russians had also bombed the Crimean canal pumping station, and he didn't mention it as a "missed opportunity" either. That too would have been a useful safety valve. It's upstream a ways, here on Google Maps, and I recently saw images of it, from Helms or whoever, damaged in another attack that's obviously disputed. 

Previously Oct 24, following UAF strike w/HIMARS, "an active discharge of water is currently going on through the canal system to protect Kherson in case of flooding when the locks of the Kakhovskaya hydroelectric power station are blown up." (https://t.me/readovkanews/45201) All other methods of release would go into the river, which was overtaxed that spring. Having these alternate drains to assist, one of them helping supply Crimea, might have really helped avoid disaster. It's probably thanks to Ukraine these two release valves were deactivated - at least one of them AFTER the lock was plugged up and just about as the dam got deactivated. 

As for the power plant's turbines, Helms says "Russian occupiers destroyed K-HPP's ability to generate electricity." Well if so, they don't admit to it. Nov 15 "Head of the occupation “administration” of the Kherson region Vladimir Saldo has told Rossiya 24 TV ... “the turbines don’t produce energy and there’s no need for it.”" He doesn't say why, but "the occupation “administration” of Nova Kakhovka in the Kherson region had fled the city due to shelling." Probably Russian shelling, huh? Russia’s occupation ‘authorities’ report that Kakhovka Hydroelectric Station’s electricity production halted — Novaya Gazeta Europe Nov 12 "The head of the Novokakhovka urban district, Vladimir Leontiev, said that the Kakhovskaya hydroelectric power station had received serious damage after numerous attacks by the Armed Forces of Ukraine. According to Leontiev, it can be restored within a year." https://t.me/readovkanews/46721

I'm not sure, but if the turbine is not operating, maybe it can't turn, and thus can't let water pass? Not a good design if so. And I'm not sure how any known attack could disable them, but I imagine there are a few direct and indirect ways the war criminals in Kyiv could, would, and did do this. This too is likely thanks to Ukraine. 

Floodgates were used, and heavily, to drain the reservoir all Spring. This was because the Russians wanted ... NOT QUITE the maximum possible volume of water? It's a good idea in that case to rotate the drainage to minimize localized erosion (or so I gather). But as Helms complains "Russian occupiers stopped active water management after 1 DEC 2022 for no apparent reason," using just the same few gates constantly. 

Helms thinks they were using just the same 2, or maybe 2-3 floodgates (favoring 2, based on 2 cranes?). Geoff Brumfiel thinks it's 3 - gates 5, 6, 7. It could be 8 is closed even with a crane parked above it - maybe opened at time, but not that I've seen. Gate 1 is leaking after an apparent, direct impact from a HIMARS missile, helping to wash out the nearby supports for the roadway that collapsed on June 1 - gate 3 seems open under the other crane in some views, including this one at right. (and BTW a Russian witness to the collapse says it started at damaged gate 1 and progressed from there. Mark Krutov on Twitter Initial damage is totally consistent with that.) 

Still, as noted, that's probably too few gates for safe operation, likely causing localized erosion that - along with a heavy and high-centered water load - led to the collapse. (other, relatively minor damage from attacks maybe played a small role as well,) But noting the role of the always-half-running gate 1, it's thanks to Ukraine, and to the US government and to ME, and every US taxpayer who helped pay for that particular HIMARS rocket. 

To the extent the operators shunned gate 2 and started at 3, they didn't add to that problem as much as they might have. As for why other gates weren't used: some on the far end couldn't be opened after November 11 because the Russians blew the top off the dam there. But many in the middle presumably worked (there were 28 in total). As noted in my part 1, snipers targeted the dam, perhaps the civilian plant staff, and the cranes probably required exposed workers to move and set them up. They were left in place, as it happens, the furthest they could get from the Ukrainian side. Ukrainian shelling might also play in, damaging the rails the cranes move on or otherwise complicating the process. (the HIMARS strike at gate 1 did also destroy the rails there, FWIW). Again, this is mainly thanks to Ukraine and their enablers. 

Overview, aside from Crimean canal pumping station also allegedly damaged by the Russians (and "2-3 gates" is incorrect, FWIW - it seems to be 4):

TAKEN opportunities: Russian Occupiers Flooding Themselves (and Don't Ask Why)

All of that might be moot anyway. In fact, the few gates used were doing fine as far as moving volume, and they were releasing about the maximum amount they safely could. Whatever exits the dam operators used, there are still limits to the amount the river can deal with, and this might have been the real choke point that prevented the dam from lessening its terrible load. It's possible to pass on too much water and flood those downstream, and it seems the dam operators avoided this just barely, and maybe not at all by the end. 

In the first post, I noted exceptionally high water level on the immediate downstream side of the dam. Here's a better view from June 2, cropped on the area of interest: (full image: File:Kakhovkabefore1.jpg - A Closer Look On Syria (shoutwiki.com)

The three little islands I noted and even most of their trees are submerged - the last few trees are visible just right of center. The higher coastlines with houses we see, on the higher north bank, are just threatened here. Parts of the south bank (Russian-held) here or downstream might already be flooded. New water is piling on top of this, spreading out, still raising the level moment-by-moment. And the reservoir is still splashing over 100% full. 

What, are the Russians trying to flood out Kherson here? The vicious war criminals! And reckless! Releasing this much water, and so narrowly, they might have eventually washed out supports for the dam. The occupying "orcs" released way too much water for their own and everyone's good! 

Ukraine's version of "vatniks" seriously said just about this. Investigator.org.ua, May 16: "Despite the statements of Gauleiter Vladimir Saldo, an uncontrolled discharge of water continues at the station, as a result of which the coast of Nova Kakhovka is flooded and the military positions of the occupiers are flooded. ...The video filmed on May 11, 2023, shows that powerful streams of water flow uncontrollably through the open and partially destroyed by the invaders of the [floodgates] of the Kakhovka HPP. You can also see the elevated water level in the Dnieper and the flooded coast in Nova Kakhovka." (emphasis mine) They say video supports Ukrainian media reports "of flooding on the left bank of the Dnieper," which had been denied by Russian-aligned officials. It probably does. Now what does the highlighted part show? Add 6/24: They also report how one Russian soldier died, drowned in a swiftly flooded dugout, and also note "At the same time, the water level in the Kakhovka reservoir decreased slightly. ... "

CIT May 17: "In the lower reaches of the Dnipro, a flood caused by a damage to a hydroelectric dam flooded Russian trenches." lol.

David helms noted these rising levels and was only concerned with military opportunities for Ukraine. May 19: "The lower Dnipro downstream of the #KakhovkaHPP is running at peak water level (about 1 meter above average). Landing spots ordinarily accessible will be flooded while other locations usually unreachable may be accessible."

So the Russians released way too much water and were flooding stuff. Also, as Helms was just fuming, the Russians didn't release nearly enough water. 4/28: "2-3 K-HPP sluice gates are open, releasing too much water in fall dry season, not enough in spring wet season." Everybody knew the reservoir was dangerously overfilled, and this was why. Those dastardly villains! How could they manage to release way too much water AND not enough water? That's just the depth of their evil, I guess? And how could they even acquire so much water with which to wreak this mindless havoc? We'll come back to that.

Hydroweb: downstream station in excess

Here, the river is usually no deeper than 0.5m or 0.6m, but we can see essentially flood levels reached well before the final dam collapse on June 6, in line with the boastful pro-Ukrainian reports at the time.

KM089 - about 4-5 km from the dam. usual high 0.5m. This level was reached by Jan. 30. With increased output, a deep 0.85m was reached by Feb. 26. About the same was reported March 25, then it was 1.32m on April 21, and finally 1.73m by May 18. That's a rise of 0.88m is less than 2 months - the river here got 3.5 times as deep as usual well before the dam ever gave way. 

KM076 a few kilometers down, the river narrows and gets a little deeper: normal high here is about 1 meter. 0.91m was reached by March 24, and 1m by April 3. (readings every 10 days here) On 4/13 it was 1.18m, on 4/23 1.3m, on 5/3 1.5m, and 5/13 1.59m, Then on 5/23 it was down a bit to 1.54, and back up to 1.6m on June 2. Level rises 0.7m from 3/24 to 6/2. 

An abnormal rises to about the same level was also recorded last May, and in December, with unusually low dips in between - all after the Russians took over, for what it's worth. KM089 has less resolution, but also shows a sustained high level last spring. The reservoir stayed within normal range then, and this might be why. I don't know if anything was up with that, or in December except - isn't December about when the reservoir got all empty? I think the Russians did THAT on purpose, sensing the spring water offensive to come, once all those escape valves were disabled in October-November, and maybe seeing high water input already starting.

KM062 further downstream: normal high 0.7m. On March 24 it was still just 0.5m, but reached 0.7m by April 20, and by May 17 it's 1.11m = 0.61m rise from 3/24 to 5/17 (last pre-collapse reading).

KM046 normal high 0.5m, reached by March 11. On April 7 it was just a bit higher, 0.45m. By May 4 it was nearly double the normal high at to 0.95, then down a bit to 0.9 on 5/31 - median 9.25 at end = ^0.425m from 3/11 to 5/31.

KM053 River Ingulets with its own flow but where it empties into the Dnieper ... normal level: ~0m to 0.58m. The Ingulets would later be flooded backwards from the Dneiper, and we probably see a bit of that already in May. It was at 0.36m on March 11 (a small decrease from February), then way up to 0.62 on 4/7 and 0.95m on May 4. Just as with the above (KM046), April to May rose a bit faster - as with above, early to late May saw a slight decline to 0.88m on the 31st (last pre-collapse reading). May median = 0.92m - rise from March to May: 0.56m.

On balance, it seems this stretch of the river rose at least 0.6m in these 2 months, and likely higher yet in the first days of June (KM098 seems to show this beginning). This is clearly because the Russian dam operators were pumping out an abnormally large amount of water, as all other evidence already suggested. 

This rise is fairly close to levels that debatably require mass evacuation. When dam worries surfaced last October, "Acting Governor of the Kherson Region Vladimir Saldo said "Wherever serious flooding is possible, the water will rise by one meter, according to calculations. We really should help people to leave the region faster." (TASS) There was probably worry over the panic this might cause, and the next day Kherson Region’s deputy governor, Kirill Stremousov (who would later be killed in a likely Ukrainian rocket strike) said: "Even if the Khakhovka dam is hit, the water level will rise by one meter, one and a half meters at the most. .... A critical situation is ruled out," he said on Russia’s TV Channel One. (TASS) Of course things had changed by May and June of this year so that a full collapse seemed possible. The recorded rises were acceptable for weeks on end, reportedly did cause harmful flooding by mid-May, and yet the reservoir remained almost 100% full.   

So this might show the Russians filling the dam until it broke, or just filling it to maximize the damage from their quiet bombing of the dam, and then ... maybe trying to minimize the damage at the last minute. Well, trying to minimize it by quite a bit, over the last 2 months ... Sure, maybe. 

But where did they manage to get all that water in the first place? Upstream, obviously. But Ukraine controls the upstream dams, and surely they wouldn't want to be the water carriers, as they say, for this Russian plot or blunder. Well it seems that, somehow, they were tricked into sending the Russians just the weapon they needed to unleash devastation on Ukraine and on themselves, to Ukraine's ultimate immediate and immense (military) advantage.

Upstream Stations in Decline

I was starting at Kyiv HPP because there's no dam further upstream that isn't in Belarus ... but then, I learned of that reported flood from a dam "failures" in Belarus. Then I didn't pin down any facts there anyway, so ... whatever excess the river held before it hit the Ukraine border, all these stations would handle that plus the rains they gather from tributaries in their stretch, and anything else the dams upstream send their way. And that sounds like it should add up to a significant amount. 

April 15 "In #Kyiv, photos are published from both banks of the #Dnipro River. In total, about 500 households in #Ukraine are flooded. The State Emergency Service reports a seasonal rise in water levels in 6 rivers. The situation is currently under control." https://twitter.com/Geoff_WarNews/status/1647232587906072576

April 18 photos of mild flooding  https://twitter.com/suspilne_news/status/1648348339019063296

April 18 Donetsk, Chernihiv, Kyiv and other regions are suffering from floods this month, which are expected to peak on April 22. SES reported “a seasonal rise” of numerous rivers in 🇺🇦 including the Dnipro, Desna, Seym, and other rivers. https://twitter.com/tvtoront/status/1648352835573018625

We should see every dam pitching in to help contain this, with a clear but safe rise in levels after mid-April. But somewhere near Kyiv: Flash News, April 20: "The water level in the Dnipro River has dropped by 10 centimeters compared to yesterday, KCSA reports. According to the KCSA, the peak of the flood is expected on April 22." They needed to make room for that.

I'm not sure where/what reservoir this refers to, but Kyiv city is at the north end of the Kaniv reservoir. That excess was sent downstream, for the other dams to hold or pass on. More excess would come, and also probably go. No one was worried about buildup here. Flash News followed up with "the expected maximum level of water harvesting on the Dnipro within the city of Kyiv does not pose a threat to the city's industrial facilities and residential areas." 

The danger would be transported south by the Ukrainian-operated dams and concentrated in one spot, possibly by some series of mistakes, or deliberately playing a massive shell game with weaponized water. But it's not such a clear picture as the dire one downstream, largely from about half the relevant reports failing to appear. Whole months pass without even the usual periodic update, where moves in such a criminal game could pass without even the kind of indirect view we have in other spots and other times. 

Kyiv HPP reservoir (north of the city):

KM0995 (on the Dnipro at the north end of Kyiv HPP reservoir): entries every 4 weeks to Jan. 12 at 103.19m (sharp rise from 12/16, 102.47), then I think 4 missing entries (Feb, Mar, Apr, early May) before a May27 entry shows a net gain to 103.49. That's a net rise 0.3m 1/12-5/27 BUT It's not remotely clear how much of the April deluge was retained vs. just passed through even between the reported dates, let alone over the crucial months of data we can't see. Coordinates: 30.5631, 51.2564 (lon, lat)

KM0981 River Teteriv, ~6km from the reservoir, likely to reflect its level somewhat - normal range 102.3-103.5m. Entries came every 10 days up to January, with a sharp rise from December, then between 1/1 and 4/20 there's no data (10 missing entries) but a net rise of 0.66m. 4/20 103.5, 4/30 103.72, 5/10 103.67 and 3 missing entries since (5/20, 5/30, 6/9). 6/19 has a reading of 103.46m - back to nromal range, down 0.26m between peak on April 30 and June 19. Total 13 entries missing (1/10, 1/20, 1/30, 2/9, 2/19, 3/1, 3/11, 3/21, 3/31, 4/10, 5/20, 5/30, 6/9) Coord: 30.2459, 51.0406 (lon, lat) 

KM0986 same river but a bit upstream, less reflective of reservoir levels but with no missing entries. Normal range: 102.5-103.4m - that was reached by 2/21 and it rose from there - 103.48 by 3/20 and 103.51 by May 13 - 2nd highest it's been since 2019. But it falls to 103.3 on 6/9 = down 0.21 5/13-6/9. According to the above, levels stayed high until at least May 27. The drop should be after this.

Kaniv reservoir: 

KM0881 south of Kyiv HPP, - reports every 4 weeks to December 31, then 2 missing for Jan/Feb before 3/22 shows 91.94m - a 0.34 rise since December. Then 2 entries for April and May are missing before June 11 shows 91.88m with abnormally high uncertainty of 0.14. That's likely because the low end is the true one, and 91.74m would make for a 0.2m drop since March. Otherwise, that wide margin means it might be down just 0.06m or even up by 0.14m. Either way, in between these entries should have been the April rains, but these were probably passed on along with that probable bit off the top of what they had before. 

As noted above, a stretch of river near Kyiv that might be part of the Kaniv reservoir shed 0.1m in a single day, April 19-20.

Kremenchuk HPP reservoir (Lake Kremenchutska):

As I gather, this usually holds close to half the water in the Dnipro cascade, with Lake Kakhovka holding most of the other half. Relevant Hydroweb stations include 2 on the Dnipro south of Kaniv HPP, one on the lake itself, and one on a side river where it joins the reservoir. Going north-to-south...

KM0741: water level normal range: 79-82m - that whole range is covered in the dramatic rise from Dec. 12 to March 3, then it goes up to 82.48m on May 6, then down to 82.05m June 2 = 0.43m drop.  Coord: 31.6901, 49.6368 (lon, lat)

KM0735 normal levels: 79.3-81.9m. March 24: 81.71m. April 20: 82.34m. May 17: 82.17m. June 13: 81.65m = 0.69m drop between peak on April 20 and June 13. Coord: 31.7716, 49.6304 (lon, lat)

Note that rivers, even when dammed, are (or can be) more volatile and variable than a reservoir/lake. Its changes can be more rapid and drastic, so a change of 0.47 0.69 at these narrow points, might equate to just 0.2 or 0.3m change across the actual reservoir.

Lake Kremenchutska itself (direct reading) is worth dwelling on. Water level normal range: 77.6-81.53m. Highest previous peaks: ~81.6m in 1993 & 94, 81.66 in 1995 - more recent years it was kept to 81.3 and then 81, then in 2020-2022 more tolerant, rising to 81.5 at several points. So 81.5-81.6 could mean basically flood levels, or undesirable anyway. 

March 22-28 is when level rise past the normal high (to 81.44m) and then tops out 1t 81.57 a few times as it took in water from upstream. But he held level drops a few times, followed by a refill: 0.24m drop on 3/31, refilled by 4/3, 0.27m decline to 4/20, refilled by 4/24 to a top level 81.57m. That's higher than normally allowed, but tolerable; it stayed close to that up to May 27, when 81.57 was again recorded. But then they just had to shed some water: it was lowered to 81.35 by June 9, with a slight refill reported later the same day. That's another 0.22m sent south at the worst possible time. Next update: 6/17 81.23m. Then 6/18 81.12m. That's another 0.23m off the top either during or just after the massive flooding. (there are 2 other dams in line that would have to pass this on for it to count as post-collapse flooding of Lake Kahkovka - otherwise it might be refilling those after they had drained themselves somewhere in between the publicized reports)

3/28-6/9 overall: down just 0.09m for a steady appearance. But with that flood and all those rains moving through, this half of the reservoir probably should have retained some of the load. Instead they passed it all on, with a moderate 0.9m off the very wide top of the huge amount they had built up ahead of the flood season. And why was that? Did they want to force a crisis situation to ensure everyone would send the maximum water down to Dnipro HPP?

Lake water volume variation - something I didn't notice before. 1/31 to 3/22 sees a steady and pretty massive filling from 0.76km3 to 3.69km3, with reports every 10 days at most (sometimes 2 in a day). That makes sense; as it widens at the top, every cm up contains more water. Every 0.22m off the top is way more than 0.22m off the bottom or middle. And that effect grows with every increase in the level.

This inflow was in preparation for the flood season? An adjustment: by March 31 it's dropped to 3.37m. With the possible Belarus flood and April rains, the volume rose just a bit, but with less views: they go monthly now (28-day cycle), and they skip May. Comparing April 20 at 3.53m to June 17 at 3.69km3, it seems they took on some water. But it's hard to say how much they SHOULD have taken on, and how much was passed through. Then a new 6/18 reading shows they're back to regular updates and some more water was just passed downstream (final volume 3.48km3)

KM0667 River Tiasmin: alongside the lake, some lazy kilometers from entering it, likely reflecting its level mixed with the Tiasmin's own rain-swollen flow. It rises from 74.38m on February 18 to 74.65 on 4/13, faster in the 2nd half from 3/17. April 13 is roughly when the heavy rains started and flooding was reported. But the level is pretty steady by May 10 at 74.67m, but with unusually high uncertainty of 0.17m (so maybe 74.84). It was down to 74.4m on June 6 (unc. low - 0.06) = drop of 0.1m, 0.27m, or 0.44m, depending, between April 13 and June 6. This seems more like reservoir levels than a flowing river. And the level falls likely more than 0.27m sometime between May 10 to June 6. According to the above, it should be after May 27. 

Kamianske HPP reservoir: 

No direct readings on the Dnieper on this reservoir nor very close to it - Helms likewise shows no level here in his thread (see image below). The following is kind of a side-note on a smaller dam I found too far up a side river to matter - except it might show additional inflow engineered back in February or March - and another river I checked for comparison.

KM058 Vorskla river - at 49°13'33.2"N 34°16'48.7"E - Google Maps over 20km north, looking a bit uphill, so not likely to reflect the reservoir level. But it's also south of a small dam with a small reservoir (just south of Poltava) - "Nyzhni Mlyny" HPP - Google Maps

Up 0.89m 2/20-3/19 - then 4/15 and 5/12 after the late-April rains, it shows a drop to 66.82 then to 66.01 by June 8. (5/12 66.82m, unc. 0.09 ignored - 6/8 66.01m w/big uncertainty of 0.25 = 65.76-66.26.) That makes a drop of at least 0.56 or perhaps up to 1.06m between May 12 and June 8. It was draining, perhaps, into an abnormally low reservoir. And maybe it was left dry during the rains because the emptied reservoir upstream was refilling.

For comparison: KM0692 River Govtva-gruz-ka: high levels ~76.76 12/20-3/11, with a February entry missing - then falling (with the April rains?) - 76.39m on April 7, 76.37m on May 4, 76.11m on May 31 - down 0.26m during May - draining, perhaps, into an abnormally low reservoir. There is no particular surge in March as seen above - because there was no upstream reservoir emptied into it then? Coord: 33.9591, 49.4795 (lon, lat) 49°28'46.2"N 33°57'32.8"E - Google Maps - it's a pretty sporadic river, seeming to be about 0m deep just meters away from this little reservoir that doesn't look 77m deep or even connected the the river. But by marshes or whatever, it seems to be continuous and flows into the Dnieper (quite a ways from here).

Add June 25: ZAES/(Rus)Energoatom on Telegram, May 10 2023 speaks of water held until then, reaching dangerously high levels, at "Sredneprovskaya hydroelectric power station" - which on Google Maps takes me to Kamiamske HPP. "IMPORTANT! At the Sredneprovskaya hydroelectric power station, the Ukrainian authorities have not discharged water from the Dnieper since April, due to the fact that it rises, flooding the settlements [upstream]. At the moment, local authorities are evacuating the civilian population, which they themselves flooded." There were confused-seeming fears that this dam would fail and then "the flow will sweep up to the Kakhovskaya HPP itself. And since the locks of the Kakhovskaya HPP are in an emergency condition, the discharge will lead to their destruction and the development of flooding events downstream." 

But in that crisis situation, Kamianske presumably sent the built-up excess on to the Dnipro HPP, sometime probably just after May 10. Dnipro may have depleted itself prior to that by getting Lake Kakhovka 100% by May 6. This massive re-charge plus late May runoff would allow them to keep the lake flooded all through May and into June.

Zaporizhzhia HPP (aka Dnipro HPP) reservoir

This is the final dam in the Ukrainian cascade, the one to finally pass on every fatal gallon that had accumulated between later March and late May. It's also the one whose shut floodgates will decide how dry the reservoir crossing will get after the flood. But they don't have much record of water levels in the relevant span.

KM0427 on the river well north of the main reservoir, or just south of Kamianske HPP. Normal level range: 51.2-51.8m. It was low in January and February, then rose 0.25m between Feb. 3 and March 2. but no data since (no readings for late March, April, May or June). Coordinates: 34.9414, 48.5069 (lon, lat) 

Late add: reports are back as of 6/18. It was 51.4m then - 0.3m drop since March 2.

KM0397 River Mokra-sura just 500m from the reservoir and clearly swelled by it/part of it. UP 0.65m 3/9-4/5 (50.98 - 51.63m) before the worst April rains. An entry for May 1 or 2 is missing. By 5/29 it's down from its high to 51.46m. That's a drop of 0.17m since April 5. Coord 48°19'46.6"N+35°07'35.8"E

May 28 rains led to flooding in Dnipro city - that too would flow into this reservoir, probably in time to be sent on and help crack the Russian dam. But it doesn't show at KM0397 on the 29th. 

KM0491 River Samara (east of Dnipro) - rise after 4/15 (rains?) then a steady high level on 5/12 and 6/8.  This could reflect the reservoir level somewhat, but at some 30km upstream, probably not. It's a very winding river that would be slow to drain April rains. That's probably all we see here, as I scrape for clues. Coord: 46.782, 33.1356

Consider this was THE final and direct input to Kahkovka lake, was much more shed after May 29? Kakhovka levels showed heavy output through May, roughly matched by input to maintain a steady near-full level, with surges to all-full notable on/by May 8 and May 20. Those might be extra injections from Dnipro. Last LK entry before collapse was June 4, and I wouldn't be surprised that it missed at least one final and sizeable injection of water.

Conclusion

So it's hard to say how much water passed through all these stations, except we can say it's what wound up downstream, and it added up to a hell of a lot, and more than the Russians could manage. From Zaporizhzhia north, flood season was wrapping up by early June, with levels relaxing and rivers draining to comfortable levels. This seems natural and normal in itself, but just a few kilometers south of there, the worst flood season ever was just about to reach a terrible climax the Russians and the locals were all but powerless to prevent.  

David Helms consider Mikael Valtersson's accusation of Ukraine flooding Lake Kakhovka (after and before the collapse),  and claimed in reply: "If Vatnikersson was correct, the dams would be below normal water level, but all were > storage level!" Based probably on the same data I've cited, he shows no problem at the 5 upstream dams by comparing the few reported levels with some historical norm, for the dates April 17, April 28, and June 19, completely skipping the more relevant details considered above. And as a meteorologist, he probably knows better. 

What's more helpful is to check what they held before the April deluge vs. after, wonder why it falls, wonder how and when, and why there are so many gaps in the bookkeeping of it all. In fact, his colored ovals show - to some extent - a swelling of the river that's shuffled around and ultimately passed south, as I have proposed. He also skips both of the final reservoirs in his consideration. I didn't find anything close enough to the Kamianske reservoir, but I'm not sure why he didn't consult KM0427 or KM0397 for some idea of the crucial lake at Dnipro HPP.

Helms: "The IG video shows Dnipro HPP discharging water. Why? Because in April the Dnipro River was in flood. That's was responsible dam operators do." He doesn't seem to know the actual levels there, but ... "Responsible dam operators" refuse to carry their share of the flood, and pass it all to the Russians, along with a bit of what they had before, as 4 allied dams upstream all did the same? "Irresponsible" dam operators ... well, they're Russian and all their equipment gets broken, and then their dams burst and - just like they wreck their own pipelines, bridges, cities, and maybe nuclear plants - the cruel, moronic, subhuman "orcs" obviously did it all to themselves. That's all you really need to know. Maybe someday they'll end up wiping themselves out in death camps.

Add 6/24: David Helms isn't just a NAFO guy on Twitter - he was cited by NYT! (Val on Twitter knew that)

"It is unclear exactly how the water level rose so significantly since then. But David Helms, a former U.S. Air Force and National Oceanic and Atmospheric Administration meteorologist who researches dams, said the Russian forces [their approved dam operators -ed] seem to have kept too few gates open to control the flow of winter snowmelt and spring rains. Likening the effect to a leaky bucket, Mr. Helms said that too much water has been entering the reservoir. "What the river is doing is dumping a lot of water in," Mr. Helms said. "And it's far exceeding the discharge rate."

More gates should have been opened in the circumstances, making downstream flooding not just bad but disastrous. But "Russian forces" didn't, perhaps couldn't. This terrible decision had been forced - and he doesn't note that it would be terrible - by the circumstances. As Helms puts it, that was down to "the river" just dumping it on this one reservoir in particular. He doesn't mention how Ukraine operated that river. But there ate the end of it, like "a leaky bucket" (??) the Russians didn't drain ENOUGH and/or allowed too much come in from the Dnipro HPP gates. There's nothing "Russian forces" could do to stop that inflow except to forcibly take over management at this other dam. Was it negligent of them to to leave Ukrainians in charge? There's a case to be made for that.

P.S. (maybe to move) Flooding After?

... well, even with the dam intact, the downstream towards Kherson was intentionally flooded to about half the level they worried required evacuation. After the dam was destroyed, in a more total way than considered last year, levels rose at least to about 6 meters above normal. 

KM076 June 11 entry: 6.81m (normal ~1m)

KM062 June 13 4.22m (normal 0.7)

KM089 June 14 entry: 4.77m. Normal 0.5. This is 8 days after the collapse. Was it getting worse? KM089 is usually shallower than 062, but it's deeper than it here, and on the next day, when levels are supposed to be receding. Is this indicative of continued flooding past what the reservoir held? 6-7 days estimated for it to drain. That would be by the 12th or 13th. Kakhovka Reservoir Water Levels Shown Before and After Dam Was Blown Up (newsweek.com)

Gerashchenko: "June 11 Water level in it fell down to 10,07 meters (it decreases about 6-7 centimeters per hour). By estimates, this decrease will continue for 6-7 days." (from now? is that an unexpected delay to the 17th or 18th? or does he mean it should be done very soon?) https://twitter.com/Gerashchenko_en/status/1667961642934583296

June 16 view of the dam area 10 days later looks pretty flooded still - but then the far side (bottom) is much lower than it was on the 6th (8m+ above the normal 0.5) - maybe 5-6m deep? - and this is about the same as in the reservoir - and the intact lower part of the dam might be close to surfacing/preventing further outflow ... Uh ... I dunno. Consider all that stuff.:

June 15 photo shows levels greatly receded near Kozats'ke (silos at 46.7798948,33.3205627) - just 4-5km downstream from the dam https://twitter.com/I_P_News/status/1669328429479776258

Hydroweb records LK: June 9 11.86m, bottoming out their chart - with a new re-scaled chart, By June 17 (I think*) it was dropped to 3.91m, then ... refilled greatly to 8.9m on the 18th. 

I'm not sure what that means. How can it fill back up? Is this just a fluke of super-low levels? I imagine a floating sensor might come down in a shallower pool, separated from the main river, and was then refloated by locally rising waters during an inexplicable post-collapse surge. For example. 

* I went back to check that page, and the graph has been reverted to how it was before those entries. I wonder why? (cropped screen-grab above is from June 20, 1:56am Pacific time)

Would it even be helpful to maintain such a flow? The initial flooding was terrible and could hardly be worsened just by keeping the same areas underwater for longer. At a certain point, Ukraine would want to close the gates, stop the flooding, let the reservoir dry up ... and then maybe cross to attack. 

Videos and satellite views show a dried up reservoir to the north around the Zaporizhzhia Nuclear Power Plant (its cooling pond is separate and remains full) - a flatter area than here at the dam. (pic forthcoming)  Erik Zimerman on Twitter: June 12 Dnipro HPP (at Zaporizhzhia - the last dam in Ukraine's line) closed up over drying river bed, and we can guess the reservoir behind that dam is not being flooded from upstream any longer.

Mathias Holmgren @HolmgrenTweets · Jun 12 "Just heard from @noclador on @MriyaReport . Ukr have closed all inflow gates upstream of the Nova Kahkovka dam basin. In a few days the basin may be close to fully emptied, and filled with dry sand. So up to 100 km wide area of easilly fordable terrain, w no RU fortifications." Those "lucky" Ukrainians. There were reports on the 15th that they had crossed the river, but further south near the dam, and had laid siege to Nova Kakhovka city, leading to clashes. " Right now, a shooting battle is going on near Nova Kakhovka. The Armed Forces of the Russian Federation, presumably, repulse an attempted landing by the Armed Forces of Ukraine. This was reported by the correspondent of "Izvestia" Emil Timashev. Shooting is also heard in Tavriysk and Sosny." (Takriysk anyway is just east of NK) https://t.me/izvestia/134317 

video https://twitter.com/Sprinter99880/status/1669446430279213057 

A drone strike on a hospital was also reported. 

https://twitter.com/Sprinter99880/status/1669676044838748163

And then there's the ZNPP. And other implications. ...

Add July 10: In this post, I didn't name who ran the dams, first from not knowing it, then from it's being awkward. But now, what did Ukrhydroenergo, who operated these 5 dams, have to say about their relentless flooding on the Kakhovka reservoir?

Reuters report, March 27 "Ihor Syrota, director general of the state-run Ukrhydroenergo hydropower generating company ... voiced concern about what would happen if water levels fell further at the Kakhovka reservoir ... The level has fallen because Russian troops who control the reservoir, and also the Kakhovka hydroelectric power station and dam, have let some water out through sluice gates, he said. ... Syrota said the level had risen since then thanks to the winter thaw" and also to Ukrhydroenergo's dams generously passing that all on to the Russians: 

"They (the Russians) are discharging a certain volume and we have raised the level to 14.30 metres from 13.50-13.60 metres. But still the gates (of the dam) are open," Syrota said. Still, with the gates just as open, the Ukrainian dams foiled Russia's possible drainage plot way too well. Through April they raised the level further and further, to and clear past the safe range, until it was 100% full by May 6. 

Along the way, they boasted on April 19 "The flood is subsides. #Ukrhydroenergo hydroelectric power stations on the #Dnipro and #Dniester regulate [water] levels, avoid flooding, and continue controlling discharges in compliance with safety standards." "In compliance with safety standards," as they say, Ukrhydroenergo kept filling the reservoir until it was 100% full, and then kept pouring so it stayed that full for one month until the dam burst. Then Ukrhydroenergo claimed it was a Russian bombing, with strangely specific allegations ready within hours.

ON THE NIGHT OF JUNE 6, RUSSIAN OCCUPATION FORCES BLEW UP THE KAKHOVKA HYDROELECTRIC STATION

"06-06-2023 - As a result of the explosion of the engine room from the inside, the Kakhovka HPP was completely destroyed. The station cannot be restored." No explanation was given for the dam's partial collapse 20 minutes before the final and mysterious event where part of the engine room might have been blown up. 

Ukrhydroenergo were keen to show the results of "Russia's" crime once it was done, but I don't see any sign of them complaining about Ukrainian shelling that damaged the dam and some of the reservoir's other release valves, and act completely unaware of this disastrous situation they engineered, that was probably the main cause of the collapse. In fact, they seem to have deliberately exclude this dam from their considerations as they pretended to manage all this water safely.

May 29 https://t.me/ukrhydroenergo/3569

The decline in water levels continues ... As a result of water harvesting through the reservoirs of the Dnieper Cascade, fluctuations in water levels were observed within the range of 1-10 cm per day, with a predominance of subsidence.

The volume of water in the cascade of the Dnipro reservoirs as of May 28 was equal to 47,616 cubic km, which is 3,768 cubic km higher than the volume of reservoirs at the normal support level (NPR), without taking into account the Kakhovsky reservoir, the volume of water in the cascade is equal to 27,116 cubic km .km (by 1,458 cubic km exceeds the volume of reservoirs at the NPR).

June 2 https://t.me/ukrhydroenergo/3591

"There is a decrease in the flow of water along the rivers. Water levels in the reservoirs of the Dnipro Cascade continue to decrease." Not at the disregard Kakhovka dam. It was nearing the end of a maxxed-out month and about to collapse.