Projects: Fuel Polishing and Filtration System

In the midst of sorting out a fuel delivery problem that would leave us without an engine at the most inopportune time, we discovered that the current fuel filter setup aboard Ariel is unacceptable. As we were approaching the channel at Pentwater, MI, after a quick cruise north from Muskegon with 20 knot winds, we decided to start up the engine earlier than we normally would because we had a sneaking suspicion that the surging we experienced earlier in the day was indicative of something serious. Sure enough. She wouldn’t catch. She was cranking fine, but she didn’t seem to be getting fuel. I went below and took a look at the engine, cracking bleed points to check for fuel. It seemed as though we had fuel, but the 5-8 foot seas and the 20-25 knot wind made working below rather difficult. By this point we’d sailed past the channel, rounded up to a beam reach, and were sailing straight out from shore, the channel directly behind us. With the sun quickly setting, and no desire to sort things out on the water, we decided to tack onto a reciprocal course and head into the channel under sail alone.

As we approached the channel, I scurried around the deck putting fenders in place—the channel is quite narrow, and it was clear that if we didn’t make it through the channel and into Pentwater Lake there wouldn’t be an opportunity to tack, much less turn around and head back out. We entered the channel at a blazing 7 knots, the wind over our starboard beam, Ariel heeled way over giving the people on the breakwater a show. Half way down the channel, the wind became incredibly fluky, eddying around a large dune to the south. Suddenly the wind seemed to be coming from 360 degrees. At one moment we were on a starboard tack heading straight down the channel; the next, the wind was coming over the port bow, forcing us toward the southern breakwater. At this point, realizing that the staysail was just blowing Ariel’s bow downwind—and closer to the seawall—my dad yelled to let the staysail sheet fly, allowing the main to drive Ariel forward so he could head her back up into the wind.

Shifting yet again, the wind now blew us sideways toward the opposite breakwater. I positioned myself at the bow, ready to fend Ariel off of the corrugated steel barrier, as my dad uncleated the main and staysail halyards, letting the sails fall to the deck. By this time, quite a crowd had gathered along the pier to watch our progress. I glanced up into their faces as Ariel slid closer and closer to the pier. Once within reach, I leaned into the pier with all my strength, letting out a horrendous grunt, quickly glancing aft to see if my dad was doing the same at the stern—he was.

Somehow, we were able to prevent 16,000 lbs. of boat from crashing into the seawall, saving Ariel from some nasty scrapes to her gelcoat and rub rail—none of the fenders was of use, both of them managing to settle into the valleys of the corrugated steel wall. A few from the crowd gathered along the wall offered to take docklines as they inquired about our predicament.

“Yeah, you have to be really careful in here,” one of them, apparently a local, said knowingly. “The wind can be very unpredictable.”

Before long, a motor boat powered alongside and asked if we needed a tow. I tossed the husband and wife duo a long line cleated to our bow, which the husband proceeded to wrap around a skimpy metal eye on his stern intended for the weight of a skier. I shook my head in disbelief, imagining the thing snapping off his boat and flying back at me as he throttled forward.

“This thing weighs 8 tons,” I yelled. “You might want to...”

He couldn’t hear me over the roar of his engine. He dropped the boat into gear and throttled ahead. The line drew tight, groaning a bit under the load.

Walking aft to collect the stern line, I noticed that Ariel’s rear port-quarter was about to slam into the seawall. In a moment of enlightenment, I grabbed the wheel and steered counter-intuitively, throwing the rudder hard to port— swinging her stern to starboard and away from the wall. With centimeters to spare, Ariel’s rear slid by the wall, narrowly avoiding a nasty scrape.

Things got exciting one last time as we prepared to anchor. Failing to recall Newtonian laws, the motorboat driver put his boat in neutral to ease the tension on the line so his wife could cast it off. Panic spread over his face as he realized that we were bearing down on him. Jamming his boat back into gear, his wife stumbled aft as the boat lurched forward, the line drawing tight again. A moment later he gave it another go, this time running aft and untying the line himself. Like the first attempt, we bore down on him. This time, however, he worked quickly with the line, made a lame attempt to toss it to me, and throttled forward just as Ariel’s bowsprit was about to pass over his transom. I waved goodbye, waited for Ariel to lose her momentum, and dropped the anchor (Chart of Pentwater Lake).

When we finally tackled the fuel issue—a day later—we discovered that Ariel was suffering from a major buildup of crud in her diesel tank. After removing the Racor spin-on element, catching it and the spilled fuel in a ziploc baggy, I gave the bag a good shake and held it up against the sunlight coming in through the companionway. It was virtually opaque; I could just make out thousands of tiny black specks held in suspension.

We were able to source a comparable replacement for our R24S (2 micron) filter from Snug Harbor Marina, an R24T (10 micron) filter, and returned to the boat. Rather than just replace the filter, bleed the engine, and be on our way, we decided that it was time to change the order of the filters to a more logical progression: Fram CP1110PL (˜10 microns), primary; Racor R24S (2 micron), secondary; engine-mounted Fram, tertiary. For some reason, the previous owner had the Racor as the primary filter, meaning that the most effective filtering took place first, leaving the other filters virtually clean. At 25.00 dollars a pop, we decided that the Racor, with its 2 micron filtering and built-in primer, ought to function as the secondary filter, saving money by filtering the major particles with the cheaper Fram element.

This arrangement would have worked well enough had it not been for the 24 years of accumulated sludge and crud at the bottom of our tank. Every time we encountered any weather to speak of, the next time we ran the engine it wasn’t long before she’d start surging—revving up two hundred rpm or so, then returning to the appropriate rpm every few minutes. We quickly learned that there was only minimal engine running time left once she started to surge and made for harbor post haste. We discovered the severity of the crud build-up during one of our many filter changes when I noticed that, despite being lower than the tank, a disconnected fuel line was not seeping fuel as expected. Thinking it odd, I grabbed the line and blew through it. To my surprise, it required a bit of force before air actually bubbled into the tank with a burst. Relieving the pressure, fuel began flowing freely from the line—something was occluding the pick-up line.

After investing over 100 dollars in filters, we finally spent several hours one evening while at anchor in Holland, MI, drawing out as much of the offending sludge as we could with our Oil Boy fluid extractor. Shining a flashlight through the Oil Boy’s clear plastic tube, we could see clumps of growth being siphoned from the tank. We continued fishing the end of the tube around the bottom of the tank as best as we could given the limited access through the fuel gauge access hole. Approximately 8 quarts later, we had run out of storage containers for the pumped fuel. Fortunately, however, it seemed as though we had gotten most of the goop out of the tank. We bled the engine—for about the 10th time in a week—fired her up, and let her idle for a while. She ran perfectly, without a hint of surging.

Once back at our home port of St. Joseph, my dad decided to do a more thorough tank cleaning. Armed with a Baja filter, a 30 gallon trash can, and a drill powered impeller pump, he transferred all of the diesel out of the tank, through the Baja filter, and into the trash can. With the tank emptied, he attempted to scrub and rinse as much of the sludge out of the tank as possible—again, this was done through a 4 inch access hole designed for the fuel gauge assembly. He also disconnected the pick-up line and discovered that sludge had accumulated at the screen filter—a first line of defense against particles. After giving that a good cleaning, he reversed the earlier pumping process, pumping the fuel out of the trash can, through the Baja filter, and back into the tank.

To date, we haven’t had more issues with fuel delivery. This experience, however, was the genesis of a new approach to fuel management and delivery. During Ariel’s winter layup, I will be draining the tank and doing another thorough cleaning to ensure a fresh start for next season. The most important part of our approach, however, is the addition of two Racor 500FG turbine filters in series, with our old Racor R24S plumbed on a manifold as backup so that we can switch to a fresh filter if the 500’s begin to show signs of restricted flow. This setup will drastically improve the engine’s reliability factor, but it is something called “fuel polishing” that will make our crud woes a thing of the past.

Unlike tractor-trailers, trawlers, and other vessels that have a high fuel turnover rate, sailboats—especially here on Lake Michigan where tides and remote, inland slips are non-existent—often have fuel that is months old. Invariably, condensation forms inside the tank, dripping into the fuel, creating a wonderful environment for growth. It is this growth—unchecked for 24 years—that caused all of our woes. Without a higher turnover rate, there is little that can be done to prevent this problem. Even biocide will not solve the problem once it presents itself; once the creatures are dead, their carcasses will still clog filters. But how to get a higher turnover rate? Fuel polishing.

By plumbing an electric fuel pump into the fuel system that circulates fuel through the filters and back into the tank, it is possible to keep old fuel clean and free from devastating blooms of crud. Completely turning over the fuel at least three times a week is the recommended practice. This is what we hope to accomplish this winter with our dual Racor 500 fuel filter installation, and a basic fuel polishing system.

(1.15.04) After last summer's nonsense and the pain of spending over $100.00 on Racor spin-on elements, I spent some time searching Ebay for a cheaper (in the long run) alternative to our current R24S filter. I was able to find two new 500FG turbine filters for 75.00 a piece with a much cheaper element--8.00 each vs. R24S at 25.00 each. Now that I've got the filters, I need to determine the best way to set up the filtration system so that it's possible to change filters while underway, and maintain a fuel polishing circuit. After a bit of musing, I came up with the sketch above. This system would utilize all but one of our current filters, so we wouldn't be wasting any pieces, plus it allows for an easy way to prime the fuel lines and filters--not to mention ensuring the liklihood of having an engine when we need it most.

The Walbro pump, located where it is, and the priming pump on the R24 allow for easy priming on both sides of the manifold (valve). After changing the filters in the 500FGs, all that is necessary to bleed the system up to the lift pump is to close the valve located just after the two 500FGs, flip on the Walbro and let it run for a minute or so. Once fuel has been drawn through the 500FGs, it's a simple process of opening the valve to direct fuel to the engine, then operating the lift pump until there's fresh fuel at the tertiary filter's bleed points--although I doubt that would even be necessary since there shouldn't be any air between the second valve and the tertiary filter. Similarly, switching the first valve to the R24, I can easily use it's built-in pump to prime it. Here's a picture of the pump I'm after, it's a Walbro 6802, 7 psi shut-off, max. capacity of 60 gph, continuous-duty pump. The pump is often available new through Ebay or through for about 100.00.

(1.25.04) Well, here's what things look like at this point. I only had a couple of hours on Friday afternoon to begin the layout, but I got the Star Board cut to size (20"x25")--we were hoping for 25x25, but the scrap didn't have enough height to get those dimensions--and began mocking-up the fuel system layout. I've already discovered that the Racors occupy far more space than I had imagined, and things are going to be pretty tight. At this point, I'm thinking I'll probaby have to relocate the freshwater pump, accumulator, and freshwater filter somewhere else--I was hoping to have room for all of that plus the fuel filter set-up on the Star Board. Unless I get a bigger piece of Star Board, however, I don't think it's going to happen. Too bad. The biggest challenge, of course, is mounting things to maximize space and yet allow easy access to the various valves and filters.

The Starboard is a scrap I picked up from a local marine store--Wolf's Marine. Even though it's scrap, it's in good shape and cost a pretty penny--Starboard is not cheap! Here's what it looked like before I cut it.

The Starboard was a cinch to cut on a band saw. I used a flapper wheel to smooth the edges after making the cuts. With the piece cut, I laid out the fuel system pieces. It was quickly apparent that my idea of joining the two 500FGs with a small piece of fuel hose was too space-consuming. I'll have to find a coupler that will allow me to mate the two filters together with a minimum of space between them.

I was able to find the necessary brass 3-way valves and hose adapters at a local hydraulic shop. The 3-way valves are very similar to the Tempo valves sold at West Marine but half the cost--12.50 apiece vs. 25.00.

(1.27.04) I spent some time last night mulling over the planned layout and ended up making a few changes. My original idea was to mount all three of the valves separately, with hoses leading from the filters to the valves, then on to the tertiary filter or fuel tank (for the fuel polishing circuit). Since space is at a premium, I decided to mount the three valves directly on the filters--a valve on both the inlet and outlet of the 500FGs, and the third on the outlet of the R24S. This idea makes far more sense; not only does it reduce the number of connections and save space, it also provides a fairly stable mount for each valve (although I am still planning on integrating some sort of bracket into the setup to ensure stability).

I found the necessary coupler to tie the two 500FGs together at the hydraulic supply shop. It makes for a slick installation and drastically reduces space requirements.

Here are a few pictures of the filter setup as it has evolved--with the valves mounted directly to the filters.

The valve located between the two 500FGs and the R24S is the manifold that determines whether fuel is directed through the primary fuel filtration circuit (500FGs) or the secondary fuel circuit (R24S).

The valve coming off of the left 500FG will direct fuel to the diesel tank (upper barb) for fuel polishing, or to the tertiary filter (lower barb) for normal engine operation.

On the R24S side, there is the inlet from the tank (on the left) and the outlet to the engine from either the R24 or the two 500FGs (on the right, the barb angled down an inlet from the 500FGs; the barb angled toward the lever, fuel outlet to the tertiary filter).

A simple combination of valves easily isolates the filters for service without interrupting engine operation. Similarly, with one valve adjustment, fuel can be drawn from the tank, filtered, and returned to the tank for polishing.

I've got to pick up four more critical connectors before the setup will be complete--I'm hoping to do that this afternoon--and I need to mount the filtration system on the Star Board after double checking clearances and space in the engine compartment.

As mentioned earlier, I'm mounting the fuel filtration system on the starboard side of the engine compartment. In addition to the fuel system, I also have to mount the accumulator, freshwater pump and filter on the same bulkhead. Since there is not enough space on the board to mount all of those systems, I've decided to mount the systems separately--filtration system as a unit, freshwater pump, accumulator and primary filter as a unit, and the secondary freshwater filter as a unit. The picture below shows our previous layout--minus, of course, the two new Racors.

(2.17.04) It's in. I still need to plumb the fuel lines and the water lines for the freshwater system - some of which will have to wait until it warms up so I can actually bend the hoses - but the job is basically done. Layout was a bit of an issue. Working in such tight quarters necessitated the purchase of a handy 90-degree drill attachment. With that, the rest of the installation was a cinch. Here are several shots of the final set-up.

You'll notice the old R24S just above the alternator in the picture, nestled in below the new Racors and the accumulator. Just forward of the R24S is the freshwater pump.

I ended up running each piece of Starboard through a jointer to get smooth, clean edges. I then routed a bevel on both sides of the board to eliminate any sharp edges that might cut into the engine insulation. This Starboard cuts like a charm and looks quite nice with all the edges finished off and smoothed over. What's not visible in the photos (actually just barely visible in the bottom left photo) are the pieces of Starboard I cut to use as feet to keep the Starboard from completely squishing the insulation. I used a hole saw to cut the pieces, then heated them up with a propane torch to mount them to the back of the boards.

(3.18.04) I ran all of the fuel line to the filters last week and have several pictures of the finished project. I ended up securing a few of the lines to the Starboard using zip-ties. I'm very pleased with the result - it looks good and seems quite functional.

The upper fuel line carries fuel to the tank for fuel polishing. The lower line carries fuel to the engine mounted filter for normal engine operation.

The primary valve for selecting between the dual Racors and the R24s is visible on the right of the photograph. The R24s is located just out of the picture at the bottom right.

Another angle of the filtration system. Notice the R24s in the background at the bottom. The valve located on the R24s directs fuel to the engine from either the dual Racors or the R24s. You can also see the switch that activates the Walbro fuel pump for polishing, located just below the fuel polishing valve.

A shot of the R24s located below the dual 500FGs. The line in the foreground is the fuel inlet. The valve to the right of the filter directs fuel from either the two 500FGs or the R24s. The gray electrical wire will be routed out of the way eventually - it powers the freshwater pump.

The R24s showing the primary valve that allows us to select between the 500FGs or the R24s.

A wide angle shot of the entire fuel setup.

Walbro Installation (3.18.04)

It’s taken far longer than a week to get back to the fuel pump installation, but it’s finally in – one of the projects I finished up over my spring vacation. I ended up mounting the fuel pump on the bulkhead just forward of the diesel tank in the port cockpit locker, where it tucks in nicely above the tank and provides the most efficient runs for the fuel line. I considered a couple of other locations in the engine compartment since I wanted to limit the amount of noise in the cabin while the pump was running during polishing, but things are already pretty tight in there with the new Racors and the spaces available would have required extra hose and complicated runs. Preferring straighter runs and fewer connections, I figured we could live with whatever noise the pump might produce during polishing – which shouldn’t be too much of a nuisance since we probably won’t even be aboard during polishing.

After running the fuel hose, I began thinking about how I wanted to wire the pump. I’d purchased a simple on-off toggle (SPST) by Cole Hersee that I wanted to locate in the engine compartment near the fuel polishing valve, but I didn’t have an easy way to mount it. As I was looking things over, trying to decide where to locate the switch, I finally decided to bite the bullet and pull the dual Racor mounting board off the bulkhead and mount the switch where/how it ought to be located: directly to – or through – the Starboard, just below the polishing valve, so that everything could be reached without completely removing the engine cover. This, unfortunately, required cutting off all of the zip-ties I’d just used to secure the fuel lines and removing the lines from the barbed fittings. I was motivated, though, by the fact that I was doing the job the way it should be done.

With the Starboard removed, I drilled a hole for the switch and confirmed that I would have to rout a recess in the back of the panel for the switch to sit in since its threaded portion wasn’t quite long enough to pass through the board. A few careful passes with my Dremel and a routing bit, and I had created enough of a recess to thread the retaining ring onto the switch. I crimped ring connectors to a couple lengths of 14 gauge wire, attached them to the switch, wrapped the connections with electrical tape to keep them from shorting out against the Mylar insulation, and relocated the panel in the compartment, reattaching all the fuel lines, etc. Voila! I was quite pleased with the installation, and the results were definitely worth the extra effort. (Far better than my original idea of mounting the switch on a plate screwed to the board)

I ran one of the wires coming off of the switch to the electrical panel where I wired it into a terminal bar (something I added at that time) that supplies a constant 12v from the house bank. The other wire ran straight to the positive lead of the fuel pump, of course, with a 5 amp fuse wired in to protect it. I also used a 6 gauge wire as a static ground that I was able to tie into some of the main grounding wires that came from Cape Dory. With that, I turned some valves and flipped the switch. The pump whirred to life and the filters filled with fuel.

The Cole Hersee toggle switch that controls the Walbro pump. An "on-off" plate from the local marine store adds a finished look.

Not the best shot of the installation, but the picture provides an idea of where/how the pump is located. Again, this is the port cockpit locker, just above the diesel tank.

I had to add a tee to the existing fuel return on the tank for the polishing circuit. The line coming in from the right is the engine fuel return, the line from the left is from the Walbro.

One thing remaining to be sorted out, however, is the presence of a pocket of air in each Racor 500FG just below the turbine. I called Racor to ask about this and was assured that it’s completely normal and doesn’t affect the pump’s performance – something I find very hard to believe. The Racor technician I spoke with told me that “you’ll even notice that there’s air in the top of the filter during operation” – again, something I find equally hard to believe. When I cracked the lid on one of the filters while the pump was running, fuel spilled over the top, belying the technician’s claim. At this point I’m thinking about unbolting the filters and tilting them sideways to get rid of the air pocket - of course doing this every time the filters are drained seems completely ridiculous. Perhaps another call to Racor, and another technician, might explain some things.

(3.23.04) Another Racor technician - or perhaps the same one, for all I know - assured me that the pockets of air are perfectly normal. He also told me that the filters tend to "run down" a bit during operation - i.e., the fuel level in the filter drops - leaving air space at the top of the filter. When I told him about my experience cracking the lid and spilling fuel, he said, "Well, sometimes they run high, and other times they run low." I guess the best I can do is to tip the filters to dislodge the two pokets under the turbines and hope it's okay. We shall see.

(9.25.04) This could have been added months ago, but I've just been far too busy. The old Perkins fired right up on launch and the air pockets present in the 500FGs haven't caused us any problems. Apparently the Racor tech knew what he was talking about - go figure. The filtration system has worked marvelously. Long before Ariel ever went in the water for the season, I had the fuel polishing circuit doing its thing while I was aboard working on other projects. During that time there was very little fuel in the tank, so the hope is that the Walbro was able to cycle enough fuel to clean out any remaining sludge. I do think, however, that to get a really good cleaning it's going to be necessary to remove the inspection plate and squirt diesel under high pressure into the tank to stir things up, all the while running the polishing circuit. Perhaps this is something we can get to soon after haul out. For anyone out there interested in a fuel polishing circuit, let me know and for a modest sum I'll install one for you! You do have to live in Michigan, however; I'm not too interested in traveling far.

(8.4.06) The fuel system continues to function perfectly. The only instance of "failure" was due to operator error. During our sail from Muskegon to Holland ('06) the engine began surging as it had during our first summer aboard ('03). I went below and switched the valve to the secondary circuit. Dad cranked the engine back up and it continued to surge until dying completely. I then realized - all too late - that I had only switched one of two valves on the filtration system, starving the engine of fuel and forcing it to suck air. After bleeding the engine completely, a miserable task at best and painfully miserable over a hot engine, we fired her up and decided that we needed to write up a SOP sheet for some of the more complicated systems aboard. So here it is, my new fuel system schematic and instructions:

back: projects