It has always bothered me that while we often use one inch throughhulls, pumps, and hoses, the native fittings on most fat sacs are 0.75 inch. I've wondered what would happen if we could decrease the flow restriction caused by those 3/4 inch fittings. I had some time this morning to run some tests on ways to do that.
We cannot change the fittings on the fat sac. However, we can add more of them and run them in parallel. Two 0.75 inch fittings have a combined cross-sectional area well in excess of a single one inch pump or hose, so if we're going to get any benefit it will show up with two fittings and 3+ won't add much.
The obvious way to do this is with a Y (or "wye") fitting. In other words, a one inch hose to a one inch Y to two short sections of one inch hose to the sac fittings.
It turns out that one inch Y's are incredibly hard to find. Those that are available are, oddly enough, made with high wall thickness which restricts their internal diameter - the exact opposite of what we're trying to achieve. What I'd like is a Y made with the same wall thickness as the Fly High W741 or W746. The best I was able to find had an ID of 0.670, worse than the sac's native fittings.
This led me to consider T's. Obviously, a T inserted in the "normal" fashion would introduce huge flow restriction due to the turbulence caused by the hard 90 degree turn toward each side. But this can be solved by plumbing the T with the straight-line flow along its long axis. This yields a straight through flow path with a "side" port.
In this configuration, one would expect virtually all of the flow to pass straight through the T as long as there are no downstream restrictions causing backpressure. That's ideal for this application, since what we're trying to do is recover losses. The flow doesn't need to be equal through the two fat sac fittings. We just want to make sure that if there are restrictions - if the pump could be delivering more water - that we provide a path for that extra pressure to do useful work.
Here's a photo of the components I used for testing:
In this photo I have the test hoses and fittings connected to the T to illustrate what I was describing above. "Normal" flow should shoot straight through the T, while the side port should benefit from backpressure in the system, if any. In this way the T itself introduces as little new restriction as possible.
When I searched for T's, my thought was to find some intended for that "funny pipe" used in sprinkler systems. However, it seems that industry has stopped making dedicated fittings and how relies on standard Schedule 40 and Schedule 80 barbed fittings. That's unfortunate, because 40 and 80 have thicker walls than is necessary. While the T's I used have larger ID's than the fat sac fittings, it's not as much as I'd hoped.
On to the tests. First, I did a flow test on my standard setup: One inch hose to a W746 to a W743 that would be screwed into the sac. (It's important to include everything the water actually sees on its way into the sac.) I used that flow rate (~0.33 gallons per second, or roughly 1200GPH from my 1600GPH-rated Rule 29B pump) as a baseline.
Next, I switched to the T setup shown in the photo above. I started with just the T and hose and no fittings on the ends of the hose at all. With this setup, basically no water came out of the side port.
I added the W746 right angles, figuring they would add some backpressure, and sure enough now we had water coming out of the side port. This proved the basic concept: That the Fly High fittings are introducing backpressure that sacrifices some of our pump capacity.
Finally, the W743's (grey threaded sac fittings on the end) were added, which is the real-world situation the water sees as it flows into the sacs. Now the flow out of the side port was significant. That makes sense, because the majority of the flow restriction in a one inch ballast system comes from Fly High's W743 and its 3/4 inch internal diameter.
So the question is: Do we really get any net increase in throughput? The answer is yes, but not as much as I'd hoped. I measured a 12.5% increase in flow rate (to ~0.37 gallons per second, or ~1350GPH) with the T setup shown above. Definitely an improvement, and a bargain considering the low cost and ease of installation (no extra pumps), but not earth-shattering.
Next I tried the Y. The measured increase there was about half that of the T. Given the wall thickness and small ID of the Y, I was mildly surprised there was any increase at all. This leads me to believe there is serious merit to this concept but at present we are being held back by fittings with excessively thick walls intended for much higher pressures than our ballast systems use.
In summary, it is indeed possible to gain increased throughput cheaply and easily by paralleling two fat sac fittings. However, I believe the increase is being held back by the wall thickness of available fittings. If we could find T's or Y's with thinner walls, the increase could hit 20%.
If I'd seen 20% today, I'd have retrofitted this concept onto both ends of both stern fat sacs this morning. But for 12.5%, it can wait until I winterize. In the meantime, the search for thin-walled, low pressure one inch barbed fittings continues.
We cannot change the fittings on the fat sac. However, we can add more of them and run them in parallel. Two 0.75 inch fittings have a combined cross-sectional area well in excess of a single one inch pump or hose, so if we're going to get any benefit it will show up with two fittings and 3+ won't add much.
The obvious way to do this is with a Y (or "wye") fitting. In other words, a one inch hose to a one inch Y to two short sections of one inch hose to the sac fittings.
It turns out that one inch Y's are incredibly hard to find. Those that are available are, oddly enough, made with high wall thickness which restricts their internal diameter - the exact opposite of what we're trying to achieve. What I'd like is a Y made with the same wall thickness as the Fly High W741 or W746. The best I was able to find had an ID of 0.670, worse than the sac's native fittings.
This led me to consider T's. Obviously, a T inserted in the "normal" fashion would introduce huge flow restriction due to the turbulence caused by the hard 90 degree turn toward each side. But this can be solved by plumbing the T with the straight-line flow along its long axis. This yields a straight through flow path with a "side" port.
In this configuration, one would expect virtually all of the flow to pass straight through the T as long as there are no downstream restrictions causing backpressure. That's ideal for this application, since what we're trying to do is recover losses. The flow doesn't need to be equal through the two fat sac fittings. We just want to make sure that if there are restrictions - if the pump could be delivering more water - that we provide a path for that extra pressure to do useful work.
Here's a photo of the components I used for testing:
In this photo I have the test hoses and fittings connected to the T to illustrate what I was describing above. "Normal" flow should shoot straight through the T, while the side port should benefit from backpressure in the system, if any. In this way the T itself introduces as little new restriction as possible.
When I searched for T's, my thought was to find some intended for that "funny pipe" used in sprinkler systems. However, it seems that industry has stopped making dedicated fittings and how relies on standard Schedule 40 and Schedule 80 barbed fittings. That's unfortunate, because 40 and 80 have thicker walls than is necessary. While the T's I used have larger ID's than the fat sac fittings, it's not as much as I'd hoped.
On to the tests. First, I did a flow test on my standard setup: One inch hose to a W746 to a W743 that would be screwed into the sac. (It's important to include everything the water actually sees on its way into the sac.) I used that flow rate (~0.33 gallons per second, or roughly 1200GPH from my 1600GPH-rated Rule 29B pump) as a baseline.
Next, I switched to the T setup shown in the photo above. I started with just the T and hose and no fittings on the ends of the hose at all. With this setup, basically no water came out of the side port.
I added the W746 right angles, figuring they would add some backpressure, and sure enough now we had water coming out of the side port. This proved the basic concept: That the Fly High fittings are introducing backpressure that sacrifices some of our pump capacity.
Finally, the W743's (grey threaded sac fittings on the end) were added, which is the real-world situation the water sees as it flows into the sacs. Now the flow out of the side port was significant. That makes sense, because the majority of the flow restriction in a one inch ballast system comes from Fly High's W743 and its 3/4 inch internal diameter.
So the question is: Do we really get any net increase in throughput? The answer is yes, but not as much as I'd hoped. I measured a 12.5% increase in flow rate (to ~0.37 gallons per second, or ~1350GPH) with the T setup shown above. Definitely an improvement, and a bargain considering the low cost and ease of installation (no extra pumps), but not earth-shattering.
Next I tried the Y. The measured increase there was about half that of the T. Given the wall thickness and small ID of the Y, I was mildly surprised there was any increase at all. This leads me to believe there is serious merit to this concept but at present we are being held back by fittings with excessively thick walls intended for much higher pressures than our ballast systems use.
In summary, it is indeed possible to gain increased throughput cheaply and easily by paralleling two fat sac fittings. However, I believe the increase is being held back by the wall thickness of available fittings. If we could find T's or Y's with thinner walls, the increase could hit 20%.
If I'd seen 20% today, I'd have retrofitted this concept onto both ends of both stern fat sacs this morning. But for 12.5%, it can wait until I winterize. In the meantime, the search for thin-walled, low pressure one inch barbed fittings continues.
Comment