Raw water flow rate for refrigeration heat exchanger?

[Microship]

Hi from s/v Nomadness in Friday Harbor...

Over the past few weeks, I've noticed an accumulation of salt water (yikes!) in the little compartment under the galley, and of course quickly checked the two local thru-hulls to be sure I was not about to sink at the dock. No problems there.

I finally crawled painfully under the stove and discovered a 2 drops/second leak at the pump used to circulate raw water through the heat exchanger for the AC refrigeration system. It's an old Shurflo 1.6 GPM fresh-water pump (8050-204-033, which has the new part number 2901-0213). Given the substantial salt build-up and corrosion around the body, I assumed that this was simply the wrong choice for the application... but Shurfo tech support was quick and friendly with email response this morning and suggested that this model is OK to use for salt water. Their more powerful Blaster would certainly be overkill.

Without the water cooling it just works less efficiently with airflow... although it's a fairly enclosed space and that doesn't work so well in the summer. So I do need to fix it, and was about to order a new 2901-0213... then realized that I have on hand an old and essentially unused Shurflo 100-000-21 that is rated at 1 GPM. This is described as a fresh-water non-flammable-liquid pump (wobble plate, I believe), and the ports are 3/8" barbs instead of the more generous 1/2" NPT on the original. But lower power, lower noise... and already in stock!

I'm about to try it, but I really have no intuition at all about how much water flow is needed for this application. I can't imagine that it would take much (especially in the Pacific Northwest, but I don't want to limit functionality to this region). Does anyone here have a sense of how much water needs to move through this assembly in order to extract heat when the cold-plate fridge is doing its thing? I think the little one, especially with barb adapters, would be about half the water flow of the original.

I don't know the model or even branding of the fridge system... there was no documentation at all, alas. I could post a photo, if that would communicate scale... it would be kind of nice to know what it is if any of y'all are experts in these things.

(My guess is that the installer just used this pump ten years ago because it was readily available... and that once it started leaking, the pump body started to corrode.)

Cheers and thanks for any insights on this!
Steve

[mcarling]

Worst possible case is that you'll have just under 56% (area of 3/8 versus 1/2) of the cooling capacity you had before. You didn't write how effective your air conditioning was before. If it was marginal, using the smaller pump might not be a good idea. If it was capable of very quickly cooling very hot air, then I would go ahead with the pump at hand.

[Microship]

This is the AC cold-plate refrigeration system (top-loading fridge and freezer), not air conditioning.

When it is running, I have a choice (DC switch) of using the water circulation or not. If I don't, it does the job with airflow over a heat exchanger and heats up the space under the galley sink unless I remember to open the door. Normally, of course, I just leave the switch on and it cycles the pump as needed.

My intuition (including touching the exit water) is that there is a LOT more water flow than necessary. I'm just wondering if anyone knows the real numbers....

Thanks,
Steve

[mcarling]

Oops! Anyway, it's the same principle. If it can quickly cool beverages or make ice, then the smaller pump would probably be ok. If it needs a long time to cool beverages or make ice, then it could take about twice as long with the smaller pump.

[donradcliffe]

How much power does the refrigeration system draw?

As an example, say your 1 gpm pump really pumped 1 gpm or 8 pounds/min or 480 lbs/hr.

Say your refrigeration was rated at 1000 BTU/hr (the 12V systems are typically 3-400). Then the temperature rise in the salt water side would be 1000/480 or a tad above 2 degrees F, which would be quite acceptable. If your refrigeration was rated at 10,000 BTU, then you would be looking at a 20 degree rise, which is not acceptable.

1 BTU heats 1 lb of water 1 degree F

[Microship]

Ah, good question - alas, I don't know the specs. Naturally, I want to dump the heat into the water instead of the cabin (though jeez, now that I think about it, why have I not been letting it help warm the boat this winter? Gotta collect some data on that... my datalogger graphs, so that should be easy).

Anyway, the cold plate system itself is a complete mystery... which at some point needs to yield to reverse-engineering. It was done well, with stainless boxes and massive insulation, pump-outs that share the shower sump pump via a manifold, remote gauges, and so on. Some installer got sloppy with an inline fuse on the DC controller side that flopped around in a hinged harness until it failed and did amazing things with a freezer full of crab bait, but the hardware itself looks robust (and is flanked by a DC fridge system). But specs? No idea, unfortunately.

Thanks for the numeric reality-check, though - that's helpful! I'll run it with the original pump, let it reach steady-state, and duct-tape a temp probe onto a boathook to see what the exit temperature is. Of course, I don't know how deep each cycle is, so I'm not sure how much that would tell me in practical terms... but it would help see if they're moving a lot of heat into the water at that rate. From sticking my finger in it last year sometime, I think not... I remember it feeling about the same as ambient seawater.

Steve

[mcarling]

Another option, in combination with the smaller pump, would be to use Aerogel insulation to improve the efficiency of your refrigerator.

[Microship]

Aerogel is awesome stuff, and I'd love a good excuse to play with it! I don't have any problems at that level here, fortunately - this was done extremely well, and the freezer is still frozen even in summer after more than 2 days without power, though ice cream is getting soft by then.

Here's the boat, by the way.

There are two fridges - AC and DC. The AC is the one I'm posting about, and the DC is a more normal off-the-shelf commercial unit that sucks too much power and isn't terribly well insulated.

I'm going to go ahead and try the smaller pump... nothing to lose, as it works with the air heat exchanger even without the water cooling (though in hot weather, that's not as effective, as I discovered while trying to keep things cold during a brutal week on the hard during a heat-wave for a bottom-job 2 years ago). The pump I'm asking about spends its life helping get most of that extracted heat off the boat, and I'm hoping a refrigeration guru will have an intuition about just how much water that takes. I don't have a sense of that myself, other than having felt only a minor temperature difference in the exit flow once, long ago, when I wasn't being very analytical.

Cheers and thanks for the thoughts!
Steve

[mcarling]

Good luck and please let us know the result.

[Richard Kollmann]

Seawater cooling flow of 110 volt or engine driven refrigeration is normally designed to produce adequate high side gas pressure. If high pressure is too high or too low peek performance will be affected. One simple way to determine if seawater through condenser is keeping high pressure at a good level is by testing or touching temperature of refrigerant line leaving seawater condenser and headed towards expansion valve. Adequate condenser cooling will keep this line warm not hot and not cold, say 100 to 115 degrees F. Most of the older systems designed by Crosby, Frigoboat, and Grunert had adjustable seawater bypasses so water flow could be controlled through condenser.

[Microship]

Ah, interesting Richard - thanks! I really need to find my way around this contraption before something goes wrong and debugging is necessary. I wrote to the contractor a couple of years (when that fuse failed) asking for info... all they sent was a bad sketch of a relay board but nothing about the machine itself. If I post a photo of the regrigeration unit and the hardware in the fridge (the one in the freezer is kind of ice-encrusted... it's time!), can you ID it? Then I'd know what to Google for more...

Cheers!
Steve

[Richard Kollmann]

Yes, I can ID unit even if home made, email several pictures and I will be able to help with any problem you have with his system. Most manufactures of this type system used standard parts that are available locally, except for the holding plates.

[Microship]

Excellent - thanks! Will do post-coffee.... much appreciated!

Steve

[Akka]

If you're using a 12-volt Danfoss BD 35 or 50F compressor, pretty standard these days, then practically any reasonable water flow will do you -- you'll not notice any difference between your 3/8 and 1/2 inch pumps. In fact, the system will probably work without any water cooling at all. As you point out, the main reason to use water cooling is to keep the locker at a reasonable temperature.

The arithmetic above about how much water you need to accept the heat generated by the compressor is, I think, correct, but it ignores the fact that you have an air-cooled condenser system which removes a fair amount of the heat (maybe all you need to remove) on its own. As long as the ambient temperature in the locker is below, say, 90 F, the fan-driven airflow should remove enough heat to allow your fridge to work. As noted above, the line going to the fridge should be warm to the touch but not hot. If you can grab hold of it for awhile, it's probably cool enough.

[Microship]

HI folks...

Richard and Akka, I found some photos in my archives... thought I'd just post them here. The cold plate is Grunert (shown is the inside of the top-loading fridge; there is also a top-loading freezer that is part of the same system). The label on the heat exchanger says Sea Freeze model SF 900r, 120V 5.5A, using 5.6 oz of R414B refrigerant. Sea Freeze is the contractor in Bellingham that did the install about 10 years ago.

I welcome any general insights, product names, gotchas, pointers to documentation, or other advice. It's been rock-solid for 4 years and I only started this thread because of the leaky pump that pushes sea-water through for cooling, but it would be great to know more about this key part of my galley. It keeps my ice cream frozen! <grin>

Oh, and Akka - thanks for the water-flow comment! I just picked up some adapters to drop that little pump in there... though it might not work if the raw water line drains back during the change-over. The Shurflo tech-support fellow just said, "The 100 Series is a positive displacement, low flow, low pressure pump for flooded inlets (no priming ability)."

Thanks,
Steve