 |
||||||||||||||||
Project: Replace Diesel Tank
|
||||||||||||||||
|
Projects:
|
||||||||||||||||
Introduction:This page is about a fuel tank replacement we did in the winter and into the summer of 2005 with an addendum on the second tank replacement in 2009. There's a lot of detailed information here and it may not be of interest to a casual reader. If you're particularly interested in Daydreamer, have at it. And if you're facing a similar problem yourself, you may well find the write-up beneficial. I know I would have loved to have found a similar discussion before I kicked off the project. There are sections on:
Background:By late September of 2004, we had been in Newcastle, New South Wales, for about 10 months. We had applied for 4 year retiree visas the previous May and were finding out that the process involved feeding very large amounts of minutia to a constantly changing set of bureaucrats and was, therefore, frustratingly long and drawn out. But no worries, while we needed to stay where we were so Immigration could contact us, the marina in Newcastle was both new and reasonably priced, it's facilities were among the best we'd ever seen, we had access to the internet and horseback riding, and the wine-growing region in the Hunter River valley was only an hour's drive away. Life was good. Until we began to start smelling a little bit of a diesel smell and seeing a little bit of a rainbow on the surface of the bilge water. And found a very. very small drip of diesel coming from an inaccessible part of our starboard diesel tank. Trouble, trouble, trouble .... starts with a "T" and that rhymes with "D" and that stands for "diesel leak"! Damn! We very quickly borrowed a transfer pump and moved as much of the diesel as we could to the port tank, then emptied the remainder into jerry jugs that went to other boats in the marina. We had stopped the drip, so the immediate crisis was over. A leak wasn't completely unexpected as we were aware of a sistership that had developed similar problems. The tanks were carbon steel and the area where we thought the leak was had been soaked by saltwater a couple of years earlier when we had a pinhole leak in one of our engine cooling hose, which is never a good combination. Unfortunately, the leak was in a completely inaccessible area. It couldn't be reached or even seen from the outside, even with mirrors. Opening the access hatch didn't help as the leak was on the other side of the baffle in the tank and too far away to reach, so I couldn't even touch the area. We flushed the tank with a water-based cleaning solution, examined the tank closely; the tank looked really good inside and we just couldn't see anything out of order. The corrosion had to be external, which is pretty much as expected. We spent a bit of time checking out possible quick & easy fixes such as internal seal coatings and putting flexible tanks inside the existing steel tanks but nothing really sounded like more than a temporary solution. For example, there was a good chance that we'd be able to seal a pinhole leak with an internal coating but it wouldn't last if the external corrosion continued, as would be likely. By this time we were expecting some friends from the US to visit so we put the tank problem on the back burner for awhile. After our friends arrived, we then made plans to rendezvous with them in Tasmania for some extended cruising and had decided to make the trip with only one fuel tank in service ... after all, it wasn't like we were going to cross an ocean or something. Once again the fates intervened when the engine decided it needed some R&R (removal and rebuilding). So we pulled the engine and handed it over to a local machine shop, took the ferry to Tassie and had a grand time for a month (ah, she's a beauty down there), and returned to the boat around the end of March, 2005. After rebuilding the engine, we tackled the tank. Demolition:So there we were, the engine was out so we could stand in the bilge, the entire inner face of the tank was exposed and accessible, but we still couldn't reach under the tank to feel the leak or fully assess the condition of the forward face or underside. What we could see/feel didn't look good so we pretty much ruled out any quick fixes. We decided to replace the tank with a new one. Of course, the old tank completely filled up the space allotted for it, so it wasn't even possible to know with any degree of confidence what the replacement tank might look like. Step one, then, was to remove the old tank. Actually, we had two tanks to remove in that a stainless steel holding tank was immediately aft of the fuel tank and the two tanks together completely spanned the distance between two bulkheads. Conceptually it was pretty straight forward -- cut away the fiberglass retaining bits, slide a tank into the center of the boat, where the engine had been, lift it up into the cabin through hatches in the sole, then out of the boat through the companionway. Good plan except that neither tank was small enough to fit through the hatch area in the sole, much less through the companionway. I spent the next few weeks in an intimate relationship with a reciprocating saw (i.e. sawzall) and a double handful of metal cutting blades. The tanks did come out, but only in pieces. Conceptual Design (Considering Alternatives):The problem at this point was to decide what the new tanks would look like and what materials would be the best choice for the replacements. I spent a lot of time on the internet researching statutory requirements and industry standards and best practices.
Configuration:The ideal end-result would be a replacement tank that completely replaced the original tank - perhaps made even slightly larger by poaching a little volume from the holding tank area. But a bit of time with a hot glue gun and some medium density fiberboard (MDF - kinda like cardboard on steroids) making up some mockups showed that there was no way a single large tank was going to fit through the companionway. Even if fabricated inside the boat, I wasn't going to be able to get it into position through the available openings in the cabin sole. I certainly didn't want to rip up the interior of the boat just so I could later reconstruct it around the new tanks.
So, multiple tanks it was to be, ones that would fit through the companionway and slip into position singly and add up to something close to the original tank volume in total.
Material Selection:A lot of marine diesel fuel tanks in Australia are made of welded stainless steel but US Coast Guard standards only allow very small stainless steel tanks, way smaller than I would want. I think this limitation stems from concerns that great care in welding procedure and material selection and use of qualified welders is required to avoid future problems with the heat affected zone (HAZ) in stainless welds. As we are a US-flagged vessel, I ruled out stainless. While we weren't legally required to meet these requirements because we weren't in commercial service, I still didn't want to be in a difficult position in some future survey.
Monel, a nickel-copper alloy, would make a great corrosion-proof tank but at a huge cost even if I could find the material. If only I could find some unobtanium ... hmmm.
That left aluminum (aluminium down under) and plastic. Aluminum has a long history as a fuel tank material and when built and installed properly, can last a very long time. But it is prone to corrosion and can fail in short order if not handled properly.
I really liked the idea of plastics, particularly High Density Polyethylene (HDPE), as they would be completely corrosion-free. There were two possible HDPE choices before me, using crosslinked HDPE that was only available as a manufactured (i.e. rotomolded) tank or as a fabricated tank, welded from HDPE sheet.
Rotomolded tanks have a very good history but are only available in certain sizes and shapes. If I had been able to find something close to the shape I needed, that would have been my clear choice. No such luck. To use rotomolded tanks, I would have either had to have customs molds built (monel tanks would probably have been cheaper!) or cobble something together out out of the tanks that were available. This last alternative would have caused a significant reduction in tank capacity and resulted in an installation that would be a plumber's nightmare.
I communicated with several firms that fabricated HDPE tanks here in Australia. Some advised strongly against trying to use fabricated sheet HDPE for a diesel tank for a variety of technical reasons. Others didn't really address the suitability of the material beyond saying they had done it before. ABYC didn't seem to like fabricated HDPE tanks for fuel service. In considering how I might fit together a series of smaller tanks, I began to worry about how HDPE would handle the inevitable point loadings and stress risers that a multiple-tank installation would cause. Ultimately these concerns, lack of endorsement by ABYC, and the advice of several vendors forced me to give up on fabricated HDPE as a fuel tank material for Daydreamer. But I kept it as the material of choice for the replacement holding tank.
And so on to final detail design.
Detail Design:It had long been my experience that unless some bit of work required unique expertise/skills/licensing/tools, I was generally able to do about as good a job as anybody I could hire, excepting maybe for timeliness, esthetics and the final finish. Having a fat ego and a thin wallet and no real worries about time, I resolved to do the tank design myself. I had been using TurboCAD, an inexpensive computer-aided drafting program, for some years. As I pondered the design I was going to be dealing with, I decided that TurboCAD just didn't have enough power to do the trick, particularly when it came to working in three dimensions. I needed to use AutoCAD, and so I did, first spend a month or two learning how to use it. This decision ultimately yielded some real benefits when it came time to fabricate the tanks.
Since the tanks were roughly wedge-shaped, knowing the shape of the top, front, and both sides pretty much defined what each sub-tank was going look like. Removing the templates and measuring their dimensions at close intervals allowed me to define their shape in AutoCAD. I could then composite all that information into the this; an isometric view of the tank space available showing three diesel sub-tanks (red-blue, blue-magenta, magenta-green), and one holding tank (green-yellow). At
this point, I now knew what t Construction of these tanks requires all sides to be cut from a sheet of aluminum and then be welded together to form the tank. The shapes to be cut out for the top, front and sides were already defined in the orthogonal view. With bit of fiddling and head scratching, I eventually figured out how to "unroll" the bottom piece, so I was able to define what it would like when flat. The details for these shapes were then transferred to a separate drawing that showed all the bits and pieces that needed to cut out for each tank, per the example on the left. I then took the detailed design
Having defined what each diesel sub-tank looked like and what all the cut-outs looked like, I was able to put together a Request for Quotation (RFQ) that I submitted to a number of fabricators in the area. The RFQ has copies of all the drawings and detailed specs for the fabrication ... if you're really interested in the gory details, you can find them here. For what it's worth, the design phase (conceptual plus final) took me about 5 months overall but I didn't put in 8 hour (or longer) days and I went down a lot of dead ends. I could probably repeat the work now in less than a month. Fabrication:AutoCAD was a very powerful design tool and being able to work in 3D allowed me to identify and fix some problems in the design that wouldn't have been apparent had I been limited to 2D. But nice as it was in design, it paid really huge benefits in the fabrication phase.
I
could've taken the same approach to provide patterns for cutting out the
aluminum but life is a lot easier than that now. There are now laser
cutters and abrasive water jet cutters that are large enough to handle a
full-sized sheet of aluminum. All I needed to do was provide the
laser-cutter operator with an AutoCAD file that showed where all the
pieces could be cut from standard aluminum stock and all the operator had
to do was define the routing of the laser cutting head and the machine
I hauled all the material off to a local welder and we spent the next three days or so welding up all the tanks. It turn out to be quite efficient as he didn't have to try to figure out what I meant by some aspect of the design drawings ... I was right there, lending a hand, and could readily identify what each piece was. True to the mock-ups and a tribute to the precision of the laser-cutter, all the pieces fit together perfectly. There were no gaps, no misalignment. Woo Hoo! The fruit of our labors, here. The pads you see on the underside of the tank serve a couple of functions. First, they keep the tank's surface away from the inside of the hull, so no water can stand against the tanks' underside. Second, the center sub-tank has a series of tabs that fit into slots in the pads welded to the forward and aft sub-tanks - this allows the tanks to lock together on a side that's otherwise inaccessible during installation. The fabrication phase took less than a week! Installation:I had previously positioned some plastic spacers on the forward bulkhead to keep the aluminum tank from resting against the plywood bulkhead (try to keep anything than can retain moisture from resting against the tank). As I brought each tank into the boat, I installed at the stainless steel hose and tubing fittings required, then slipped them into place. The tanks were bolted together, then slid into position after hooking up the fill and vent lines, and screwed to the forward bulkhead using lag bolts. A half-bulkhead, with plastic spacers, was bolted to the aft face of the tank and then glassed into place along with several hold-downs on the inner face. Connecting supply and return lines completed the installation.
Holding Tank:I took a similar approach with the new holding tank. First, a full detail design using AutoCAD, then make an MDF mockup from full-size prints of all the panels. Check the mock-up for fit, then turn over the drawings and the mock-up to a fabricator that worked in HDPE. A week later and the holding tank was in place and lag-bolted to the half-bulkhead at the aft end of the fuel tank and to the aft engine room bulkhead.
The port tank developed a similar pinhole leak during the winter of 2008/2009, so we cut our winter's cruise short and hauled out at the Marina Singlar, Guaymas in mid-March. The demolition/removal, patterning, and design process for the replacement port tank(s) was very similar to the above but we had some procedural variation beyond that point.
I was able to locate a firm in Phoenix that could use an AutoCAD file to control/cut the patterns I needed from aluminum plate but they used an abrasive water-jet cutter rather than a laser cutter. This process seemed to work as well as the laser jet, although the water jet controlling software had a bit of difficulty with a few of the AutoCAD features, requiring slight mods to the design. All in all, no big problem and the quality of the cut patterns was as good as those done by laser. The material and water-jet cutting costs were only slightly higher than I experienced in Australia.
Welding turned out to be much more problematic. The welder I used for the first set of tanks used a "TIG" welding process where there's a tungsten electrode and the weld filler material comes from a separate hand-held welding rod. The welder I used for the port tank(s) used a "MIG" process, aka wire feed, where the weld filler material comes in the form of a wire spool that feeds through a handle that controls the speed that the wire feed through to the weld and there is no separate electrode.
We started with the aft tank, the simplest one, and found perhaps 20 leaks when we went to pressure test the tank. This contrasts poorly with the first set of tanks, where we had only one leak in all three tanks. The other two tanks had similar leakage problems - perhaps a bit fewer but still excessive. The weld up time for the three tanks was similar to my experience in Australia, about 3 days, but the repairs required to fix all the leaks effectively doubled the welder's time. Since the welder's hourly rate in the US was substantially higher than the rate in Australia and the fabrication time more than double (because of the repairs needed), the total welding cost for the port tank(s) was more than four times what I had paid in Australia. And the welds looked a lot crappier too!
Lessons learned: If my experience is any guide, it appears that TIG welding is a far superior to MIG for this kind of work. And after the aft tank of the port-side set had so many leaks, I should have pulled the work and gone to another welder or, at least, come to some kind of understanding that repairs would be done at a lower (or no) cost after the initial weld up. Alternatively, some kind of hard-money quote might be the answer but quality control might become an issue.
|
