I'll start by saying that I have built a lot of plywood tanks over the years, with very few problems and certainly no major ones. The very few leaks I have experienced were fairly straightforward to locate and correct. All tanks were 24" in depth or less, and all were simple boxes, screwed and glued together, waterproofed with various types of epoxy coatings. All had one or more windows cut out, with glass siliconed in place after the wood had been epoxied. No fibreglass, no external or internal bracing (other than around and across the top opening).
But...a couple years back I built another tank utilizing the same method, and while doing so I realized that I had enough scraps and remainders of wood, epoxy, glass, etc. to build myself a smaller tank for "free", so to speak. It was designed for a specific spot in my basement, with non-standard dimensions of 48 x 19 x 22 inches high. I decided to try something different with this job...and it was a complete and utter failure.
I built the box as usual, braced the top and cut out the window opening. Now, anyone who has done any of these builds knows that you always want to consider the logistics of the build and do it as efficiently as possible; dividing the gallon of epoxy into the appropriately-sized portions for each individual coat, timing the coats, waiting for the epoxy to cure before applying silicone, etc. In my case, the logistics made it very attractive to try siliconing the glass to the bare wood before applying the epoxy; bear in mind this tank was being built simultaneously as the larger "main" project, so I was trying to just fit it into the work schedule as best I could.
Conversations with a product rep at GE seemed to indicate that the silicone would work this way; a couple of preliminary tests done in the past seemed to confirm it as well. So I bravely forged ahead and siliconed the glass to the bare wood interior. Looked, to my eye...well, weird. I had a couple inches all the way around the opening where glass overlapped the wood, and I tried to keep any silicone from squeezing out around the glass. I carefully removed what little did get out, and then applied the epoxy. The tank being as small as it was, it was awkward to paint the underside of the top bracing and, to a lesser extent, the glass/wood interface. I put three careful coats on the complete interior, and painted right over the edge of the glass and somewhat onto the interior face of it. I then laid a thick bead of silicone all around the entire glass/wood interface. That was a huge mistake.
The tank held water perfectly for over a year. Then, a very slow leak developed that was intermittent in character. After a water change, I would occasionally notice water soaked into the wood; the exterior of the entire tank was left unfinished so this was easy to spot. The leak was never fast enough that there were any actual drips or drops, the entire stand remained dry except for one corner and the no water got to the floor. But, obviously, the wood structure would eventually fail. The water would dry out on its own...and then, weeks or months later, it would be damp again. Lowering the water level sometimes seemed to help, but since the leak came and went even before I tried that, I couldn't be sure that a lower level was actually influencing it.
A couple weeks ago the fish in that tank went outside for the summer, so I slowly lowered the water level to see what would happen. Even with only a few inches in the tank, there was still a very, very slow leak, so the problem was obviously getting worse. I'd had enough; I drained the tank and started looking. Turns out that the epoxy I used didn't adhere very well to glass; wherever the epoxy coating on the interior glass extended past the smoothed-out silicone bead, it had slowly started to peel away, allowing water to find a path between the epoxy and the glass, under the silicone, all the way to the bare unfinished wood to which the glass was siliconed.
I have no excuse for this; probably the single biggest, most stupid blunder I have ever faced in the hobby. Obviously, I should have painted the epoxy only onto the interior wood, getting as close as possible to but not actually touching the glass. Then, applying that final silicone bead would have given me a complete and perfect seal, covering whatever tiny amount of bare wood remained at the interface. This seems so obvious in hindsight that I hesitated to even admit my mistake here, but I figured that it might help save someone else a problem in a future build.
On the positive side, I did learn that silicone on bare wood seems doable as a method of installing the glass. I plan on testing this further, on another build. Just too much trouble to remove this glass, then remove all the silicone down to bare wood neatly, without messing up the edge of the epoxy coating. Easier to start from scratch, but...you know...not do it like an idiot this time.
The leaky tank? Hmmm...I have been toying with the idea of getting another snake. Haven't owned one for years, but I have already convinced my wife that I should have one as a means of ensuring that my granddaughters don't grow up afraid of them. The tank is grossly overbuilt for the task, but...it's just sitting here now...
But...a couple years back I built another tank utilizing the same method, and while doing so I realized that I had enough scraps and remainders of wood, epoxy, glass, etc. to build myself a smaller tank for "free", so to speak. It was designed for a specific spot in my basement, with non-standard dimensions of 48 x 19 x 22 inches high. I decided to try something different with this job...and it was a complete and utter failure.
I built the box as usual, braced the top and cut out the window opening. Now, anyone who has done any of these builds knows that you always want to consider the logistics of the build and do it as efficiently as possible; dividing the gallon of epoxy into the appropriately-sized portions for each individual coat, timing the coats, waiting for the epoxy to cure before applying silicone, etc. In my case, the logistics made it very attractive to try siliconing the glass to the bare wood before applying the epoxy; bear in mind this tank was being built simultaneously as the larger "main" project, so I was trying to just fit it into the work schedule as best I could.
Conversations with a product rep at GE seemed to indicate that the silicone would work this way; a couple of preliminary tests done in the past seemed to confirm it as well. So I bravely forged ahead and siliconed the glass to the bare wood interior. Looked, to my eye...well, weird. I had a couple inches all the way around the opening where glass overlapped the wood, and I tried to keep any silicone from squeezing out around the glass. I carefully removed what little did get out, and then applied the epoxy. The tank being as small as it was, it was awkward to paint the underside of the top bracing and, to a lesser extent, the glass/wood interface. I put three careful coats on the complete interior, and painted right over the edge of the glass and somewhat onto the interior face of it. I then laid a thick bead of silicone all around the entire glass/wood interface. That was a huge mistake.
The tank held water perfectly for over a year. Then, a very slow leak developed that was intermittent in character. After a water change, I would occasionally notice water soaked into the wood; the exterior of the entire tank was left unfinished so this was easy to spot. The leak was never fast enough that there were any actual drips or drops, the entire stand remained dry except for one corner and the no water got to the floor. But, obviously, the wood structure would eventually fail. The water would dry out on its own...and then, weeks or months later, it would be damp again. Lowering the water level sometimes seemed to help, but since the leak came and went even before I tried that, I couldn't be sure that a lower level was actually influencing it.
A couple weeks ago the fish in that tank went outside for the summer, so I slowly lowered the water level to see what would happen. Even with only a few inches in the tank, there was still a very, very slow leak, so the problem was obviously getting worse. I'd had enough; I drained the tank and started looking. Turns out that the epoxy I used didn't adhere very well to glass; wherever the epoxy coating on the interior glass extended past the smoothed-out silicone bead, it had slowly started to peel away, allowing water to find a path between the epoxy and the glass, under the silicone, all the way to the bare unfinished wood to which the glass was siliconed.
I have no excuse for this; probably the single biggest, most stupid blunder I have ever faced in the hobby. Obviously, I should have painted the epoxy only onto the interior wood, getting as close as possible to but not actually touching the glass. Then, applying that final silicone bead would have given me a complete and perfect seal, covering whatever tiny amount of bare wood remained at the interface. This seems so obvious in hindsight that I hesitated to even admit my mistake here, but I figured that it might help save someone else a problem in a future build.
On the positive side, I did learn that silicone on bare wood seems doable as a method of installing the glass. I plan on testing this further, on another build. Just too much trouble to remove this glass, then remove all the silicone down to bare wood neatly, without messing up the edge of the epoxy coating. Easier to start from scratch, but...you know...not do it like an idiot this time.
The leaky tank? Hmmm...I have been toying with the idea of getting another snake. Haven't owned one for years, but I have already convinced my wife that I should have one as a means of ensuring that my granddaughters don't grow up afraid of them. The tank is grossly overbuilt for the task, but...it's just sitting here now...
