Chousei Coal Mine is located in Ube City, Yamaguchi Prefecture, Japan, and is part of the Ube Coal Field.
Ube Coal Field is a group of submarine coal mines that are no longer in operation.
In the late 19th century, mining operations in the Ube coal field began in earnest with the opening of a coal bureau by the Yamaguchi clan, and British engineer Maurice was invited from Nagasaki to operate the mines with the introduction of Western-style coal mining technology.
After the Meiji Restoration, management of the Ube coal mines was transferred to the private sector in 1872, and a number of coal mines were opened during this period as demand increased due to the Sino-Japanese War.
The mining of the seabed was introduced in the Ube Coalfield after 1912, and it is said that the amount of coal produced in the Ube Coalfield doubled after that time.
The Ube coalfield did not continue to operate peacefully, and several major seawater inflows occurred.
Although operations continued to be affected by the increase and decrease in demand due to the war, all coal mines in Ube City closed in 1967 due to the gradual shift to oil as the main energy source after World War II.
The coal mine where the seawater inflow accident occurred could not be rescued or the bodies salvaged, and the entire mine shaft was abandoned.
Therefore, the bodies and relics left in the submerged mine shafts were left untouched.
The chousei coal mine is one of the Ube coal mines that experienced a flooding accident in 1942 that resulted in the deaths of 183 people.
Note that 137 of the 183 victims of the Chousei Coal Mine accident were from the Korean peninsula.
It is reported that there were more Korean workers in the Ube Coal Mine than in other areas of Ube coal field.
It is estimated that about 40% of the total workers in the Ube coalfield were foreign workers, and about 85% of these foreign workers were from the Korean Peninsula.
The existence of coercion against foreign workers and the quality of the work environment are disputed in some cases, but we will not discuss those issues here.
In any case, the fact remains that 183 bodies were left in the submerged coal mine.
Piers are the exhaust and drainage stacks of the tunnels.
The original entrance to the Chousei Coal Mine has been reclaimed and we are attempting to enter the underwater mine from the pier.
Of course, we also have in mind to dig back the original entrance.
The piers protrude about 5 to 7 meters above sea level, and the water surface inside the piers is approximately the same height as sea level, but seems to be slightly higher than sea level.
We began our research and exploration of the site in July 2024.
During the initial exploration, we tried to enter the pier of offshore side (called offshore pier), but we were unable to enter the pier due to high waves caused by a typhoon a few days before the exploration date, which prevented us from working to build scaffolding to enter the inside of the pier.
During the second exploration, we were able to enter the offshore pier of the two piers.
When we went underwater, the transparency deteriorated to about 50 cm from a depth of about 7 meters.
The water inside the pier appeared to be almost entirely freshwater.
Diving through the interior structures, we found a pile of broken pipes and scaffolding at a depth of around 27 meters.
The temperature at the bottom of the water was 18 degrees Celsius, and since the seawater temperature was 22 degrees Celsius at this time, there appeared to be little or no seawater in the mix.
According to the drawings, the entrance to the tunnel was located at that depth, so we searched along the inner wall of the pier but could not find the entrance due to the piles of objects.
The next exploration will be conducted in the shore pier where the drawings show no such piping, etc.
If we are still unable to enter the tunnel, there is a possibility that salvage work of the broken piping will be conducted.
We attempted to excavate the entrance to the mine, which had been reclaimed at the time of the accident 80+ years ago.
Fortunately, drawings and other documents from the period when the mine was in operation remain, and the general location of the mine entrance can be determined.
We began excavation work with heavy equipment on September 24.
We were unable to excavate the entrance to the tunnel on the day we began construction, but fortunately again, we were able to find the entrance to the reclaimed tunnel the following day, September 25, 2024.
The cost of these excavations was funded by crowdfunding.
Water trapped for more than 80 years began to overflow from the entrance of the mine.
The frame of the tunnel was made of pine wood, and the entrance section was 2.2 meters wide and 1.6 meters high.
The water surface rises and falls with the ebb and flow, indicating that there is some kind of traffic with the ocean. However, upon confirmation, the water was freshwater, there are no large holes in the sea.
We spent several days there, clearing the ground and taking certain measures to prevent the surrounding soil from sloughing off. We also paved the pathway to the entrance of the mine with concrete.
After the land clearing, a memorial ceremony was held by the civic orgnization on October 26, 2024 with the families and relatives of the victims of the collapsed coal mine.
We conducted our second dive exploration at the end of October.
On October 29, we conducted a dive from the pier closer to shore(called shore pier), and on October 30, from the entrance of the mine.
In the previous exploration we used an open circuit for simplicity, but this time we used a closed circuit to reduce the risk of collapse because of the high possibility of getting more inside the mine.
This is because it is assumed that the breathing bubbles will contribute to the collapse, and in the long term, the oxygen in the exhaled air will degrade the tunnel frames.
According to a exploration conducted in the 1980s, there is a history of being able to descend to the very bottom because there was little piping or other obstructions inside the shore pier, so there was hope that it would be possible to enter the interior of the mine from shore pier in addition to the entrance of the mine.
We prepared our equipment on the morning of the 29th and attempted to dive to pier in the afternoon.
The pier on shore is about 6 meters high, and we had hooked rope ladders to the outer and inner walls in advance so that we could ascend and descend.
The waves were high that day, and it took some time to get to the shore pier, but we reached them safely and succeeded in getting inside the pier.
When inside the shore pier, the water was murkier than offshore pier and did not provide much of a field of view.
The visibility was clear enough to see one's hand down to a depth of about 5 meters, but below that depth, the visibility was 10 cm or less.
We descended to a depth of about 23 m, but could not reach the very bottom because of obstructions by broken steel pipes and other internal structures.
We groped around for a route, but we decided that it would be difficult to continue the exploration from the shore pier and decided to cut off the dive after about 10 minutes.
On October 30, a exploration was conducted from the entrance to the main tunnel which was thought to offer the best possibility of access to the inside of the mine.
A large number of press gathered to interview the exploration on October 30th, I began the dive surrounded by more reporters than ever before.
I had expected the water to be a bit clearer as I went deeper, but the clarity was about 10 cm all the way.
Although there were only a few moments when there was enough visibility to see own hands, I were able to get in to the point where I used up one of our reels.
Using a different reel, I continued on a little further, and although it looked like I could still go further, I decided to turn back this time because I were about to reach our scheduled dive time.
The depth at which I decided to turn back was 26.1 meters, and I was able to almost reach the very bottom which is estimated to be about 30 meters.
In addition, the water temperature was about 18 degrees Celsius, all fresh water, and the slope angle was about 10-15 degrees.
Since I think the skeletal remains are located near the bottom area, I believe that there is a high possibility that I will be able to actually recover the remains next time.
From January 31 to February 2, 2025 we carried out our third exploration.
During this exploration, it was hoped that we would actually find remains.
Although the exploration was conducted during the season of winter, with a sprinkling of snow, the enthusiasm of the press seemed to overpower the winter weather.
The diving began at the entrance of the tunnel under the expectation of the local civil society, the bereaved families, and the press, and broke through the previous point without any problem and proceeded further forward, but then a problem occurred.
At the far end of the tunnel, at the depth of 30m, the timber looked as if a jungle gym had collapsed, and it was impossible to break through that point.
On the first day, I was forced to return to the entrance, and on the second and third days, I tried to break through that point by hand with zero visibility, but in the end I could only assume that the tunnel was causing a collapse in that location.
At this moment, Yoshitaka Isaji, the leader of the diving team, began to think that it might be difficult to reach the far end of the tunnel by diving from the tunnel entrance.
As a result of this expedition, I brought back several man-made objects from inside the tunnel. The items in this photo appear to be wooden equipment used for ventilation inside the tunnel.
This was the fourth exploration between the 1st and 4th of April, although the previous exploration had ended with harsh results.
This time was conducted with Kyungsoo Kim and SU Eun Kim, friends of project leader Yoshitaka Isaji (me), a South Korean diver who runs Tech Korea.
The objectives of this expedition were 1. to see if it was possible to go further from the point currently reached by another person, and 2. to see if there was a route between the entrance of the tunnel and the collapse point to move to a side road.
The route to move to the side road did not exist on official maps near the entrance to the mine, but was said to exist according to the testimony of those who worked there, and we needed to be sure anyway.
On the first and second days, I checked whether it was possible to pass through the collapse point, but visibility was poor and we were unable to do so.
This time, we also conducted surveys using MNemo, which allowed us to clarify how far we had progressed in our previous explorations.
On the third day, we carefully checked whether there was a route to move to the side passage between the entrance of the tunnel and the collapse point, but we could not confirm its existence. We had no choice but to conclude that it was impossible to dive into the far area from the original entrance, although there was a slight possibility.
During this expedition, we attempted to use a wired muddy water camera to capture clearer video footage, but the water was so murky that even with the muddy water camera, we couldn't see anything.
During the previous exploration, the underwater exploration leader, Yoshitaka Isaji, had concluded that diving from Pier rather than from the original entrance of the coal mine was the best course of action, and had begun the task of removing the lumber and steel that blocked the passage to the bottom of Offshore pier.
Although we were unable to remove all obstacles inside Pier before this expedition, on the fourth day we dived inside Pier to check its internal structure.
I and civil group have set the date for my next exploration in mid-June and have decided to remove the obstacles inside the Offshore pier before then.
As mentioned previously, steel pipes and wood are piled up inside the Pier, blocking the way forward.
However, it is expected that there is a path leading to the tunnel at the bottom of the pier (on the map, it was thought that the old tunnel along the main tunnel could be accessed from the Pier), and it was considered that removing those obstacles would allow access to the tunnel.
The removal work was mainly carried out by hand by the local diving shop (VOX plus) and collaborated divers.
We cut through the obstacles with an underwater reciprocating saw and continued to pull them up. Obviously, steel is extremely heavy, and the wood, which had absorbed water, was heavier than we could have imagined.
Some steel materials could not be lifted by hand, and accidents occurred during the lifting operation in which a steel pipe fell from above. Therefore, it was decided to use a crane barge to lift some of the steel materials.
The work on this crane barge cost 2 million yen (approximately US$14,000) per day, and due to budget constraints, we were only able to use the crane barge for one day. Nevertheless, this work enabled us to continue removing obstacles manually.
I know I'm being repetitive, but all the costs for the exploration and removal work so far have been covered by crowdfunding and donations.
The removal of obstacles took more than a month, but finally, a sloping shaft was found at a depth of 32 meters.
The diver in charge of removing the internal structures slipped through the jungle gym-like overlapping structures and discovered a side tunnel leading to the mine shaft.
Despite the risk of falling structures above, the divers successfully tied a rope from the surface to the 32-meter entrance.
The diver in charge of the removal work advanced about 10 meters into the inclined shaft and confirmed that it was not filled in, that it continued further, and the general structure of the entire shaft.
The diver decided that it was too dangerous to continue on his own, and the exploration team took over the search.
This meant that the expedition scheduled for mid-June could now be looked forward to with great anticipation.
We welcomed the days of exploration on June 18 and 19 with high expectations.
Now, we may finally be able to successfully enter the mine passage.
The diving depth was set at 35 meters, which was deeper than expected,and although it may seem excessive, we prepared trimix for diluent for safety reasons.
We prepared oxygen at the 6 meter point on the decent rope inside the pier, and during these operations, there was a problem in which the support diver dropped the reel, but we decided to continue using a spare reel.
The lower part of the Pier was about 10cm transparent, but we successfully reached the inclined shaft by groping our way, passed through the restriction between the end of the inclined shaft and the room-like space, then reached what appeared to be an old mine passage beyond the room-like space.
Once we entered the old passage, visibility improved a little, and we could see about 50 cm to 1 meter ahead.
The old mine tunnel ran in the direction shown on the map, so we decided to follow the passage further into the mine.
Visibility of one meter was the best we could get in this coal mine, and we quickly advanced about 120 meters and use up our reel and turn back.
It was very regrettable that I lost my reel as soon as I started diving, but we secured a route into the tunnel and confirmed that it continued on, raising our expectations for the next day.
On the second day of exploration, we easily reached the point we had reached last time (of course visibility was zero up to that point), connected the explore reel to the line end, and began to advance further.
In order to limit myself, we only brought a reel that was 200 meters long, because the total length, including 100m already extended in the previous dive, is approximately 300 meters, but this was calculated as the maximum distance that could be safely returned with only two tanks in the event of a bailout.
When we passed the point we had reached last time, visibility was about one meter, but after advancing about 50 meters, there was a moment when visibility dramatically improved.
The old mine shaft gradually slopes downward, reaching a depth of 40 meters, below which is a layer of saltwater (to be precise, the fact is that the salinity is higher than at shallower depths, so it is unclear whether it is completely seawater or not).
The tunnel reaches a depth of 42 meters and then becomes completely horizontal, with the interior of the tunnel almost completely filled with seawater.
The saltwater layer has a visibility of more than 5 meters, and no floating objects could be seen in the water.
We moved forward, feeling both happy that the visibility had improved and concerned that the water was 10 meters deeper than we had expected.
As visibility improved, a gate-like structure quickly came into view.
We attempted to pass through it several times, but the passage was too narrow, so we decided to remove one tank to pass through the gate since we were concerned that forcing our way through would damage the structure and cause it to collapse.
There was a thin pillar on the inside of the gate, we don't want to stimulating it, but we tied a line to it since there was no other option, and we moved forward.
Due to the high transparency, the exploration progressed at an unprecedented speed, and a new gate appeared, but again, we broke through the narrow passageway.
As we continued further, a fork in the road appeared.
A fork in the road! We felt a sense of accomplishment at having successfully entered the inner part of the coal mine. It is truly regrettable that the battery of the action camera used for filming ran out just before reaching this point.
The path split into two, and I decided to check the right path first.
Swimming far down the right passage, I came across a brick wall, but it ended in a dead end after about 70 meters.
Upon further reflection, I realized that this passageway was a dead end on the floor map, so I decided to proceed in the opposite direction. Perhaps I was a little nervous, which is why I didn't notice it at first.
I started moving toward the opposite side and advanced about 20 meters, but my TTS (Time to Surface) reached about 50 minutes.
Δ+5 is about +8, so I decided that if I returned now, the decompression time would be about 100 minutes, and decided to return at this point.
The total dive time was likely to be around 200 minutes, and since we had said that the planned dive time would be around 120 minutes, we thought it would not be a good idea to extend it any further. In any case, there was not much reel left, so it would not have made much difference.
This dive resulted in a total dive time of approximately 200 minutes, with 100 minutes at the bottom and 100 minutes of decompression time.
After finishing the dive, we left Pier and headed to the beach to report the results to the media.
I think we can finally say that “we were able to enter the mine” this time.
For our next expedition, we plan to install bailout tanks in the mine shaft, and then carry out a plan with a bottom time of 150 minutes and a decompression time of 215 minutes.
The day when we will finally be able to recover the remains may be approaching. We eagerly await our next expedition.
Based on the outcomes of the previous exploration, we conducted this expedition in two phases.
First, we carried out equipment installation and preparations to ensure safety during July 30–31 and August 6–8, 2025. Then, from August 25–27, we planned to invite two Korean divers—Mr. Kyungsoo Kim and Ms. Su Eun Kim—who had also dived with us during the February exploration, and proceed with the search for skeletal remains and survey measurements.
In particular, for safety management, we anticipated long decompression (approximately 100–150 minutes) at depths shallower than about 6 meters. To make decompression procedures smooth and reliable, we newly established a “decompression station.”
This station was designed to hold spare tanks and equipment, allow divers to maintain stable depth, and be adjustable so it could be moved to any depth between 6 meters and the surface. This enabled both safe and practical operation—such as adjusting decompression depth according to the tide and using it for surface preparation as well.
Tasks to be completed before the Korean team’s arrival (July 30–31 and August 6–8, 2025)
Tasks to be conducted together with the Korean team (August 25–27, 2025)
We canceled the dives planned for this period because high waves—caused by southerly winds from an approaching typhoon—made it difficult to approach Pier by boat.
Meanwhile, setting up additional bailout tanks and decompression equipment, preparing installation work, and filling various gases required significant time—ultimately about three full days. In that sense, not being able to dive during this period may have actually been beneficial for completing preparations without omissions.
After the typhoon’s influence subsided, conditions finally allowed diving.
Before my dives, tanks were sunk in the following configuration to prepare for tasks and emergency response.
At 6 meters depth
At 21 meters depth
At the bottom of Pier (entrance to the passage)
I added QC6 connectors and inflator hoses to the regulators on the installed oxygen tanks so they could be used not only as an independent external breathing gas, but also connected to both the diluent and oxygen systems of a rebreather.
Because the bailout tanks are strictly for emergencies, I planned to leave them in place for a long period until the exploration is completed. To prevent corrosion of the aluminum tanks, I attached magnesium rods as sacrificial anodes and used galvanic corrosion effects to reduce damage to the aluminum.
To reiterate, all of these installation tanks and associated equipment were funded entirely through crowdfunding and donations led by the citizen group. I recognize this initiative has both supporters and critics, but it is notable that it attracted significant public interest and made preparations on this scale possible.
Originally, I planned to focus only on transporting tanks and recording. However, the day before the dive, the citizen group strongly requested that I attempt to reach the main passage. I therefore changed the plan at the last minute.
On August 6, I transported emergency tanks and filming equipment up to the restriction (a narrow section that is difficult to pass), along with two reels loaded with thicker line for use in a passage with many sharp structures. I left the equipment—including the camera gear—underwater, and planned on August 8 to recover part of it while aiming for the deeper section.
For filming, I used an OLYMPUS OM-D E-M1 Mark II with a Nauticam underwater housing, two RGBlue lights fixed on tripods, and also an INON large underwater tripod. I considered various ways to transport this large amount of equipment, and ultimately divided it into two mesh bags and carried them in our arms.
Transporting the heavy and numerous items was difficult, but I successfully made three round trips from the bottom of Pier to deeper inside the passage and carried the equipment to just before the restriction. The decompression station functioned well during ascent, allowing us to complete decompression in a stable environment.
However, challenges remained regarding coordination with the support team during depth adjustment and while waiting near the surface. The support side needed to descend each time to make adjustments, so a more efficient operational method will need to be considered.
Gate at the passage from the old tunnel to the main tunnel
As the last dive before the Korean team’s arrival, I entered the main passage and conducted filming on August 8, following the revised plan.
Immediately after starting the dive, the drysuit was damaged by many rusted metal fragments remaining inside Peya, causing a leak. I considered aborting the dive at that point, but judged it was feasible to continue for the following reasons:
Based on these considerations, I concluded there was no fatal safety risk and continued the dive.
After that, I progressed smoothly, recovered the tanks, lights, and camera that had been left during the previous dive, and advanced further. On the way from the old passage toward the main passage, I discovered a large gate-like structure—an artifact strongly reminiscent of the mine’s operating period and of high historical value.
After filming the gate and passing through its opening, I reached a T-junction. Based on the operational-era map I had reviewed in advance, I judged this to be the main passage.
From the T-junction, it was visible that the left side (toward the beach) had collapsed, while the right side could not be seen from the junction—suggesting the passage continued. I decided to try the right side (toward further inside), but after swimming about 10 meters, I unfortunately saw that this side had also collapsed further ahead.
After fixing and securing the line on the further right side, I attempted to confirm the left-side collapse point. However, visibility rapidly deteriorated due to line work and percolation (sediment falling caused by bubbles). With the late-August exploration approaching, I decided to turn back at this stage.
On the return, I gradually felt my body temperature dropping, so I promptly returned to the bottom of Pier and began decompression. At 21 meters, I ran out of the battery for the heating system. To request the delivery of a spare battery, I rang a communication rope with a bell attached. However, the support team did not recognize this as an “emergency signal,” and we were only able to meet about 40 minutes later—after I had ascended to 6 meters and some additional time had passed.
After confirming at the surface, it became clear that while they noticed the bell sound, they did not interpret it as an emergency. This was a major issue, and we considered it as follows:
Going forward, we need to set a simple, immediately actionable rule—for example, “ringing multiple times indicates an emergency.” At the same time, we must update the entire team so that each member can judge for themselves that an emergency may be occurring and act immediately.
Starting August 25, we had planned dives jointly with the two Korean divers (Kyungsoo Kim and Su Eun Kim), as described above. However, just before that—on August 20—I injured my left arm and became unable to participate in the dives. I considered canceling the exploration itself, but decided that they would proceed without me for the following reasons:
However, we partially revised the schedule: dives would be conducted on August 25 and 26, and August 27 would be a rest day.
After the two korean divers arrived in Ube the day before the dive, I provided detailed explanations of the exploration results to date, the newly installed equipment inside Pier, the tank placement status, and the procedures for each task. We then proceeded with gear preparations for the actual exploration.
During preparations, however, a diver working inside Peya reported that “a steel pipe has slipped and fallen on the route from the bottom of Pier to the inclined shaft.” Although the opinion was that “it is barely passable,” if a similar event occurred during a dive, it would pose a serious risk of blocking the ascent route.
This incident once again highlighted the reality that a large amount of steel pipes and wood remains unstable inside Peya. This risk was “recognized, but could not be fully mitigated due to budget constraints,” and the slip event is something that can occur again in the future—underscoring both the limitations and challenges created by budget constraints.
Metal film formed in one and a half days
We defined the objectives of the August 25 dive as follows:
After starting the dive, they struggled to pass the slipped steel pipe that had been a concern, but ultimately succeeded in passing through safely.
They then checked the tanks that had been sunk during the previous dive and found that the magnesium rods had corroded more than expected.
In addition, considering that during the previous exploration a metallic film (rainbow-colored film) formed on lights and plastic reels that had been installed for only about a day and a half, we inferred that the water inside the passage contains a high concentration of metal ions.
It is particularly noteworthy that metallic film was deposited even on plastic equipment. This suggests not merely aluminum oxide film, but the possibility that metal components such as iron and manganese, or metal components including sulfate ions derived from acid mine drainage (AMD), are dissolving and depositing in the water.
In general, the likelihood of acute poisoning is considered low, and even if a small amount were accidentally ingested or if diving continued, the likelihood of chronic toxicity is also considered low. However, repeated exposure to metal-contaminated water raises concerns not only about poisoning but also about irritant contact dermatitis (and in fact, we experienced skin irritation during this survey). Therefore, we judged it desirable to include a rest period after a certain number of dives.
After a dive lasting around 200 minutes, cheers rose when the divers surfaced.
They had found skeletal remains inside the passage and successfully brought them up.
The three recovered bones were blackened by the coal mine water, but showed little damage.
While “happy” may not be an appropriate word in this situation, I felt deep emotion at having achieved the first step of “finding remains and bringing back the first bone,” at seeing efforts spanning more than 30 years finally rewarded, and at believing that we may have helped turn a tragic accident toward a slightly brighter conclusion.
Based on prior arrangements, the remains were handed over to the Ube Police immediately after returning, and submitted for examination at the Forensic Science Laboratory. This follows standard procedures to confirm whether the discovered bones are human, rule out criminality, and then conduct identification inquiries.
Remains recovered from the Chosei coal mine: femur, humerus, radius
On the second day, August 26, we continued recovery of remains from the previous day and conducted exploration measurements that had not yet been completed. During this process, we also carried out additional filming inside the passage.
The photographed remains appeared as if they still retained the form from the time of death, to the extent that it felt more appropriate to describe them as “bodies” rather than “remains.” During this work, we newly discovered a human skull and successfully recovered it.
There was no doubt that the remains found the previous day were human bones, but the discovery of a skull was particularly shocking. It forced us to confront the fact that a “human being” was lying submerged underwater, and as people involved in this exploration, we could not avoid strong emotions.
According to examinations by the police Forensic Science Laboratory, all bones were confirmed to be human. The three bones discovered on August 25 were identified as humerus, femur, and radius; the remains recovered on August 26 were a skull and teeth (excluding the mandible).
I plan the next exploration for around late January to early February 2026, inviting multiple divers from overseas. I hope that Japanese divers will also gain experience diving in such a harsh environment for the future and potentially become team members later, but at present, it is difficult to realize their participation.
Planned participants are as follows (A→Z). Names in parentheses are English names:
Mr. Por and Mr. Victor are divers who previously accompanied me on underwater cave exploration on South Daito Island.
Mr. Mikko Paasi is also known as one of the divers who contributed to the rescue as a member of the international rescue team during the 2018 Tham Luang cave incident in Thailand, when 13 people were trapped.
In this way, I feel very encouraged that we have been able to build an international community where we can invite trusted colleagues from around the world and undertake underwater exploration together. At the same time, I also strongly feel that in order to remain “someone who is called upon when needed,” it is essential that I continue constant training and preparation myself.
Although the specific tasks for the next exploration have not yet been finalized, at minimum, I absolutely want to recover the remains that were discovered and confirmed this time. I will continue the exploration going forward, and I appreciate your ongoing support.
Copyright © 2024 Japan Karst & Underwater Exploration Project