on a boat

    Galleon     Contact me.

This page updated:
December 2013



From Seven Seas Cruising Association's 2000 Equipment Survey:
"Eliminate the fridge and you don't need a generator, you run the engine two hours less, take off three solar panels, cut your battery bank in half, and mooch ice cubes off everyone in the anchorage."

A freezer is even more demanding than a refrigerator,
from "Refrigeration for Pleasureboats" by Nigel Calder:
  • Takes more space, because needs more insulation.
  • Takes about twice the energy of a refrigerator.
  • Food in freezer is more sensitive to temperature rise, so:
    • Breakdown quickly causes major food loss.
    • Can't leave boat unattended / unpowered for very long.

If no refrigeration (not even an icebox):
  • No meat (except salted, canned or smoked).
  • No cold drinks.
  • No ice cream.
  • No milk (except canned or powdered or UHT).
  • Some vegetables (lettuce, celery, etc) will perish very quickly.
Maybe this is tolerable.


Main types of refrigeration:
  • Compression (aka "Evaporative")
    Used in home and most boat refrigerators.
    Uses [DC-electric-powered or AC-electric-powered or boat-engine-powered] compressor to circulate the refrigerant, which may be Freon.
    Lubricating oil mixed into refrigerant.
    • Evaporator: liquid refrigerant passes through expansion valve or capillary tube, expands and absorbs heat from inside area, changes from liquid to gas.
    • Compressor: compresses gas refrigerant, heating it further.
    • Condenser: refrigerant transfers heat to outside, changes from gas to liquid.

  • Absorption
    Uses gas heat to circulate the refrigerant, which may be ammonia-hydrogen-water mixture.
    No moving parts.
    All contents at same pressure.
    Sensitive to direction of gravity.
    • Generator: heats water-ammonia, vaporizing ammonia.
    • Condenser: ammonia transfers heat to outside air, changes from gas to liquid.
    • Evaporator: hydrogen mixes with ammonia, absorbs heat from inside area, changes from liquid to gas.
    • Absorber: water absorbs ammonia, releases hydrogen.

  • Peltier (thermoelectric)
    Used in small self-cooling ice-chests.
    Uses electricity to drive a bi-metal element; one side of element gets hot while other side gets cool.
    No moving parts.
    Very inefficient, and can drop temperature only 40 degrees below ambient.
    Can be used to heat as well as cool.
    Peltier Device Information Directory (Thermoelectric Cooler/Heater/Generator Modules)
    Askeland kits for installing Peltier air-cooled or water-cooled in a box

  • Evaporative (non-mechanical)
    Zeer pot (AKA pot-in-pot):
    From Dennis Craven on SlashFood:
    "It's basically two large [unglazed] earthen pots, one nested in the other. The space is filled with sand and water is added. A damp cloth covers the top [maybe with a clay lid on the inner pot]. As the water evaporates [through the surface of the outer pot], the inner pot containing the perishables is kept cool -- water evaporation draws heat from the inner vessel. Water is added twice a day."
    Requires a dry (low-humidity) environment with good air-flow.
    From BobRainbow on SlashFood:
    "One problem I found in the developed world was that our pots don't seem to be porous enough. They are fired too high and don't allow enough evaporation. You also need quite a high ambient temperature to make them work. I tried some while it was coolish and they did absolutely nothing."

    From MJ on SlashFood:
    "On hunting trips we had a low-tech way of keeping food cool. We had a tin box, possibly six or so cubic feet, that was covered with two or three wraps of burlap (old potato sacks). The box was filled with food (and beer of course) and placed at the edge of a creek. The burlap surrounding the box was well soaked. The end of the long piece of burlap that surrounded the box was placed in the creek so the box would be continually kept wet through capillary action. If the weather was extremely hot, every couple of hours, the burlap on the box would be soaked with water. The evaporation of the water keeps the food good for several days and the beer cold enough until you didn't care if the beer was cold or hot. (The evaporation of water from the sacks surrounding the tin box keeps the contents cooler than the temperature of the creek.)"
    Wonder if this would work with salt water, hanging such a box off the side of a boat ? Requires a dry (low-humidity) environment with good air-flow ?

  • Passive
    Use a nearby body of earth or water that is at a lower temperature than the air. Unfortunately, tropical or southern water is too warm for this.

Holding plates, and cycling:
Each type of refrigeration can be used with or without a "holding plate" (aka "cold plate") that is taken to a low temperature and then absorbs heat from the contents of the refrigerator over the next day or so.

If there is no holding plate, the system must be "constant-cycling" (with a thermostat, running often, such as 3 minutes out of every 15) to maintain the temperature. Obviously constant-cycling doesn't make sense for engine-driven or genset-AC-powered systems.

Rough rule of thumb: holding-plate system will take about 1/2 as much energy as a constant-cycling system. But a holding-plate system is more expensive and complex and doesn't run automatically.

It seems to me that the type of refrigeration should be matched to the type of power source you will use. A high-power short-duration power source such as an engine or generator would work well with holding-plate refrigeration. A low-power long-duration power source such as solar panels or wind generator would work well with constant-cycling refrigeration. Dock AC power would work with either type of refrigeration.

Heat removal:
Each type of refrigeration can use air-cooling or water-cooling, or both, to transfer heat from the outside of the system to the environment.

Water-cooling can be done either by pumping seawater through a heat-exchanger in the condensor, or by a keel cooler (metal outside the hull, connected to the condensor).

Air-cooling is simpler and doesn't require through-hulls, but is less efficient (water conducts heat better than air, and the sea usually is cooler than the air near the refrigerator).

Better to oversize your system (and pay a little extra), than to undersize it (and have to run it all the time).

Rough rules of thumb: in tropics, refrigerator load is 600 BTU's per cubic foot, and freezer is double that.

Power sources:
  • Engine:
    Good: efficient (no power-conversion losses); high capacity.
    Bad: increases engine hours; runs engine at low load; have to run engine when docked; rubber hoses to engine-mounted compressor are slightly porous, so moisture can get in.

  • AC:
    Good: mass-market constant-cycling equipment is cheap; simple hookup when docked.
    Bad: need genset when cruising; holding-plate equipment is expensive.

  • DC:
    Good: very flexible (power from batteries, wind, solar, alternator, and/or battery charger).
    Bad: probably have to upgrade whole DC system; less efficient if power-conversion losses. Many "marine AC/DC" refrigerators are expensive because they contain an inverter and an AC-powered compressor.

Compressor types:
  • Hermetic:
    Compressor and motor are sealed into a container and bathed in oil.
    Good: low cost; compact; long life.
    Bad: limited power, designed for constant-cycling.

  • Open / belt-driven:
    Good: higher power; can connect to many power sources (AC motor, DC motor, engine, genset).
    Bad: more expensive.

Other parts:
  • Filter / dryer / receiver:
    Filter protects compressor from debris.
    Dryer removes water.
    Receiver keeps expansion valve supplied with liquid, and can hold entire refrigerant charge during servicing.

  • Accumulator:
    Reservoir for liquid, to prevent liquid from getting to compressor.

DC-battery-based system may require upgrading whole DC system,
from "Refrigeration for Pleasureboats" by Nigel Calder:
  • Bigger batteries, to satisfy the power requirements.
  • Higher-capacity alternator, to recharge batteries in reasonable amount of time.
  • Fancier voltage regulator on the alternator, to charge batteries fast but avoid overcharging.
  • Maybe solar panels or a wind-generator, to reduce engine run time.

From Richard Kollmann on Cruising World message board 12/2004:
Costs of maintaining a boat's refrigerator are skyrocketing due to recent changes in the refrigerant industry and the lack of knowledge when servicing them. The ban on manufacturing R12 refrigerant has resulted in mistakes in how to service these units, costing many boaters hundreds of dollars.

Here are some things boaters should know:

Refrigerant 101: Refrigeration is the process of removing the heat from the evaporator in the refrigerated box and disposing of it to an area outside the box. The compressor circulates the heat to the condenser where it is transferred to another medium: air or seawater. It does not matter what type of a mechanical refrigerator you have; this is the basic principle of how it functions.

Refrigerant 102: Most older boat refrigeration systems used Freon R12 refrigerant, except for low-temperature freezers which used R502. Both of these refrigerants can no longer be manufactured as they affect the ozone layer and global warming. If you determine that your unit needs R12, it is best to stay with that refrigerant. Yes, there is plenty of R12 out there but it just costs more. R12 is currently selling for $20 per small can on Ebay. Remember the final authorization to use alternative refrigerants or oil should come from the compressor manufacturer and no one else. The most-used 12/24 volt refrigerator compressor is the Danfoss. It is used by Adler Barbour and at least twelve other companies. The older Danfoss BD 2.5 compressors use Freon R12 refrigerant and mineral refrigerant oil. Refrigerant 134a was not approved for these compressors because of its inability to mix with mineral oil, and to change the oil to a compatible oil might destroy the insulation on the motor wiring. There is one blended refrigerant, 409a, that can be used as a replacement for R12 in BD2.5 compressors, but only in an emergency in cool climates. Warning: never mix refrigerants, and always vacuum out the old before inserting a substitute.

Hermetically sealed compressors that are powered by alternating current (110 or 220 volt) may be filled with Alkylbenzene oil, which is compatible with HCFC component refrigerants. Blended refrigerant 409a is recommended as a replacement for R12 in systems where the compressor contains alylbenzene oil.

Externally-driven compressors, engine-driven or those with large 12/24 volt motors, are being converted from Freon R12 to 134a refrigerant. This also includes changing to a compatible Poly Ester conversion oil. If you have one of these systems, contact the system manufacturer for their correct procedures before making this conversion.

Don't forget to clean the condenser coil annually on air/water cooled refrigerators when the system performance is in question.

Nondestructive testing to determine refrigerant quantity is important to avoid contamination. You don't need a mechanic to check out your boat's 12-volt refrigerator before your cruise. Turn the refrigerator on and after 30 minutes there should be frost covering the evaporator. Adler Barbour and other similar ice box conversion units supply a three-foot piece of insulated tubing to use on the lines just outside the box. In moderate temperatures and humidity three feet of insulation should be enough to prevent moisture on the lines outside the box. In hot humid climates, it is a good idea to insulate both tubes together in a single insulation tube outside the box all the way to the condensing unit. If the evaporator is not 100% covered with frost, the unit is low on refrigerant. It frost extends outside the box on the line back to the compressor then there is too much refrigerant in the system.

Large refrigeration systems in boats generally have holding plates inside the refrigerated box that may take more than an hour to freeze. A touch of a finger on different parts of the system can tell if the system is working OK. On the line just before the plate is an expansion valve; by touching the outlet of this valve performance of the system can be determined. After about ten minutes the temperature of the outlet of the valve should be below +10 degrees F. If you place a finger on the valve outlet at this time your finger will almost stick to the valve indicating the system is probably OK. Systems with expansion valves generally have a sight-glass in the pressure liquid line after the condenser that will indicate refrigerant quantity, but you must know what it looks like when the system has the correct amount of refrigerant in it.

To extend the life and obtain the best performance from your boat's refrigerator:
  • Use the correct refrigerant and oil.
  • Only connect servicing gauges when it is proven necessary and purge the gauge set before connecting it.
  • One ounce plus or minus of refrigerant will effect the performance of any capillary expansion device system like the Adler Barbour.
  • On the new BD35 and BD50 compressor systems the actual compressor run time or cycling is no longer a true indication of performance.

I provide a great deal of free boat refrigeration information on my web site so be sure to check out the Forum and FAQ section. It you have one of the new efficient 12/24 volt compressors, the Slide Show will be of interest also.

From Richard Kollmann on Cruising World message board 4/2005:
Always look at the small-print specifications. "The above ratings are based on cruising in areas where the seawater is not warmer than 72 degrees F and the box must have medium size insulation."


New Refrigeration Systems are not always more efficient.

The ASU is a great enhancement to these refrigeration units where the compressor selected has a capacity somewhere in the mid range of the box's daily average heat load. The ASU's normal function is much like a car's automatic transmission by slowing the compressor down when high power is not needed. Slowing the compressor speed cuts current draw as much as half. Another advantage of the Auto Speed Up ( ASU ) is to override the controls when charging current is available to the battery bus.

There is extensive testing of these speed controllers in my 12/24 Volt refrigeration manual with computer analysis profiles. The results of the Isotherm Model 4201 Magnum with ASU Showed poor results because the tropical heat load on the box was to much for the range of the system and ASU. On the other end, if the compressor at slow speed is too large for the box heat load excessive current will be used.

If the system is to be used in warmer waters than the manufacturer's specification list, their figures must be adjusted to a new capacity and current draw. I use a simple rule: add 4 percent for each degree of warmer water, or subtract 2 percent for each degree F of cooler seawater.

From refrigeration article by Nigel Calder in 1/2004 issue of Sail magazine:
  • Used to recommend DC-powered holding-plate systems, but now favors DC-powered constant-cycling in many cases, because small compressors are getting more efficient and higher-output, and insulation is getting better. Also, constant-cycling units use hermetic compressors (fewer leaks), capillary tubes (no moving parts), smaller footprint inside the box, weigh less.

  • Constant-cycling AC-powered: good if usually connected to shore power.

  • Holding-plate engine-powered: most efficient (little power-conversion loss), but requires a lot of engine run-time, often at low-load. And have to run engine even if connected to shore power. Can't run unattended.

  • DC-powered: even if charged from engine, can have less engine run-time than engine-powered alternative, because batteries can be charged faster than holding plate can be cooled.

  • DC-powered is most flexible: can power from engine/alternator, solar, wind, shore power (charger), etc.

  • "Hybrid" systems such as Glacier Bay's Micro HPS are appearing. A small-holding-plate system with many of the advantages of constant-cycling systems (hermetic compressor, compact unit) and holding-plate systems (more efficient, high capacity).

  • Nice to have two DC-powered constant-cycling units: still cheaper than a holding-plate system, get to separate freezer and refrigerator, and get redundancy.

Additional from article in Mar/Apr 2008 issue of Ocean Navigator magazine:
  • Engine-driven compressor can be large and powerful, minimizing holding-plate freeze-down time and engine running time. But this is limited by rate at which heat can be withdrawn from a holding plate, which is function of surface area of evaporator tubing and temperature differential. Probably requires multiple holding plates.

  • Note that the slower the compressor's operating speed, the more efficient it is (i.e. the more BTUs of heat removed per watt-hour of energy consumed). There are significant benefits to be had from running a compressor for longer hours at slower speeds, rather than for shorter hours at higher speeds.

  • In recent years, the relative efficiency of the small compressors used in constant-cycling units versus the larger versions used in holding-plate systems has improved. Efficiency of small compressors can be further boosted by variable speed controllers (which run for longer hours at lower loads; doesn't pay if a water pump is used for cooling).

From John Dunsmoor [circa 2001]:
The most successful of the cruisers in my experience are using dual-coil holding plate refrigeration with dual compressors. One being engine-driven and the other electric, either 110 volt or 12 volt. The choice in electric is usually based on whether the vessel will be berthed a good deal of the time or has a generator in suite. I would say now a days most are going for 12-volt mainly because the new 12-volt compressors are getting a lot better than they used to be. You will find that most vessels in the forty-foot range do not have a generator, so they are using their engine to top off batteries for an hour or so anyway. The engine-driven compressor has a great deal more power than an electric compressor and will bring a holding plate down to temperature very quickly.

From Dave Richardson on the WorldCruising mailing list:
  • Don't oversize your freezer or you will be cooling a very large volume that is used less than you think. In our cruising we eat more dry, sealed and fresh than frozen. Leave more room for beer and less for meat. There is nothing like a cold beer on a hot day to get you through the clearance procedures in the islands.

  • As the material in the freezer is removed, replace it with plastic jugs filled with water. They will then freeze and provide solid, highly cold absorptive material that will balance the running of the compressor cycles.

  • On the drains at the bottom, install shut off valves [or a cork] that prevent cold from spilling through that little hole into your bilge.

  • If your boat is even a few years old, start all over on the insulation. The old material will likely be undersized, deteriorating and can be improved with modern technology anyway. This is not a fun job. But most (all) production boats use only a couple of inches of low efficiency (R2 per inch, and if it's water-soaked R0) foam insulation. Spending some time, energy and $$ on improving this with high-efficiency closed-cell insulation at R20 to R50 per inch will yield tremendous results.

  • If it has a front-opening door, get rid of it and install top access only. Every time you open the front door, the cold spills out and the seals will leak anyway. You may have to stretch to get to the bottom, but that is viewed as an opportunity to clean the accumulated grunge out anyway.

  • Check out Glacier Bay. They cut and distribute a high performance (read also high cost) insulation material, sell a great ready to install freezer and fridge lid, and then of course have AC and DC cooling systems.

  • Go cruising prepared with your own set of gauges (make sure they are for your fittings), silver solder (to repair lines), and a can of the appropriate refrigerant. You will at some point lose the refrigerant and rather than feed the sharks indigestible frozen meat it may be you are able to repair it.

  • Go with a water-cooled [instead of air-cooled] system [condenser]. It is about 10X more efficient in thermal transfer and doesn't warm up your cabin.

  • Choice of systems depends as much on your fridge and freezers volume, insulation, your battery bank, power sources etc as anything. At least in my experience, and where we spend most of our time, the system of choice is either Adler-Barbour, Frigomatic, Sea-Frost, or Glacier Bay.

  • Locate refrigerator/freezer away from heat sources such as oven and engine.

From Mr. Appliance:
  • Brush and vacuum the condenser at least twice a year for maximum energy efficiency.

  • Place two thermometers in your refrigerator, one in the freezer and one in the fresh food compartment. Check the thermometers each time you open the door. The freezer temperature should be less than 10°F and the fresh food compartment should be between 36 and 38°F.

  • When a refrigerator is warming up and your frozen food starts thawing out, most of the time the problem is not the compressor or loss of Freon but rather in the controls.

From Brian Woloshin on Cruising World message board:
12-volt-powered refrigeration gives you the most flexibility. In port you can use a battery charger. You can use solar or wind and if the sun don't shine you can run the engine.

From "Second Thoughts" article by Tim Murphy in 6/2004 issue of Cruising World magazine:
From Rob Dubin: for a DC refrigerator, get a smaller compressor motor rather than a bigger one. The batteries are much happier putting out small current for long periods rather than large current for short periods.

From Larry KN4IM on newsgroup:
[A friend's] compressor is mounted on the boat with a belt to the engine. I think this stresses the engine mounts and pulls the engine out of alignment with the shaft because I can see it pulling the engine sideways, loosening the belt. I wouldn't install one in this manner. Insist they bracket the compressor properly to the ENGINE not the BOAT.

From Peter Hendrick:
Among things that we specifically avoided based upon 1st-year experience:
Engine-driven refrigerator (we were tired of being a slave to the boat; now we have a small and very efficient 12 volt / air-cooled system that came with boat which uses 55 amp-hrs/day).

From Lee Haefele on The Live-Aboard List 7/2005:
Air-cooling versus water-cooling:

I attended a seminar on refrigeration at last year's SSCA Gam. The water-cooling did not perform better until compressor was more than 1/6 HP. Until then, the pump consumed more than the gain. THEN, the instructor advised to DO ANYTHING POSSIBLE TO NOT SALTWATER COOL. He has had many early failures due to corrosion, all total losses. Several things that improve performance: Using a cold plate instead of an evaporator, improves efficiency due to keeping refrigerant at 15F instead of colder, major difference in energy use. Variable speed compressor, smart, voltage-sensitive controls that cool extra when engine or battery charger is on. The keel-cooler in a drain seemed like a good idea, but, it may not cool well in a totally calm anchorage; also, be sure to change its zinc.

What I want but no one seems to make:
I want a 12 VDC, air-cooled, fridge/freezer, constant-cycling, drop-in unit (stock, one piece, like my current Nova Kool). But using that vacuum-panel insulation that has the incredibly high R-value.

No one makes a unit like that, as far as I can tell.

Why not ? Maybe because it is cheaper to get the increased efficiency by some other route, such as changing to water-cooled instead of using expensive insulation ? Or some other reason ?

Terrific book: "Refrigeration for Pleasureboats" by Nigel Calder.
Theory, building icebox, choosing system, sizing and building cold plate, building condenser, installing, troubleshooting, servicing, etc. Pretty heavy going in places.

BoatU.S.'s "Choosing The Right Marine Refrigerator For Your Boat"
How Stuff Works's "How Refrigerators Work"
SailNet - Curt Epperson's "Managing the Cold"
Article by G. Kevin Alston in issue 1999 #4 of DIY Boat Owner magazine
Kollmann Marine's "Insulation"
SailNet - Joy Smith's "The Art of Ice-ing"
SailNet - Sue & Larry's "Installing 12-Volt Refrigeration"
Article by Joe Minick in 3/2004 issue of Cruising World magazine
Article by Steve D'Antonio in May/June 2004 issue of Ocean Navigator magazine
Article by Nigel Calder in Mar/Apr 2008 issue of Ocean Navigator magazine
Scott Fratcher's "Trouble shooting marine refrigeration and marine air"


Nova Kool
TropiKool (Stirling compressor, CO2 refrigerant)
Minus 40

From Cecil Smith on the Morgan mailing list:
Good source of parts or service: Indian River RV Sales and Service in Sebastian, FL.

From Steve Strand on the WorldCruising mailing list:
We also lived for ten and cruised for several years with a Dometic LPG fridge and loved it, except on passages where we were heeled over. After about 24 hours on a single tack it just stopped working. I thought about a gimbal, but it just got too large and tricky. Works great except for that and the fact that you need to vent some heat. [And you need to buy LPG.]

Paraphrased from "The Boatowners Mechanical and Electrical Manual" by Nigel Calder:
  • Refrigerator compressor: want low-pressure and high-temperature cutout switches on compressor body.

  • Reciprocal compressor (York, Tecumseh) better than swash-plate compressor (Sankyo, Sanden).

From Mitch on the SailNet liveaboard-list:
First, I'm in favor of sailing with refrigeration and margaritas! Some sailor types are also minimalists, not in favor of refers (but, they usually like the margaritas).

I installed a SeaFrost system on my Tartan 37. It included both an engine driven compressor and a 12V compressor which ran from 110VAC via an inverter. It had a glycol solution which froze a holding metal block "plate" about the size of a block of ice. Two ice cube trays sat above the block and I could make ice cubes in 45 minutes. Very cool. The block would freeze meat, or whatever in one partitioned part of the ice box and refrigerated the other half of the box via holes in the partition. I could use 110VAC when at the dock, or engine when cruising underway or at anchor or 12VDC when necessary.

I would recommend something similar. Glacier Bay also has a great system. You want holding plates and a freezer section that can drive a compressor from your engine via direct or a large 12VDC motor or 110VAC. ALL of the cruisers that we met with only 12VDC refers that ran from the batteries only, without holding plates - ran their batteries dead all the time. The goal is to freeze the plates in 45 minutes or less. That way the engine/genset can do the job without running too long and make a days worth of cold, re-charge the batts and hopefully make your daily water usage (if you have a water maker).

Systems that just "trickle" all day to make cold, or electric, or water suck if you really seek cruising independence.

From David Romasco on the SailNet liveaboard-list:
Our current boat had a Grunert engine-driven holding plate system, which I've removed in favor of a super-efficient 12 VDC system using a keel cooler from Frigoboat; VERY satisfied to date.

Two thoughts: Do you REALLY want a refrigeration system that relies on running that very expensive diesel at extended idle to make it work? And if you are looking at 12 VDC systems, remember that almost all manufacturers build their system around the same Danfoss compressors.

From GATOR / Calypso on Latitudes and Attitudes Cruisers Forum:
Our last three "cruising" boats had different systems:

Alberg: Ice Box.
Always plenty of ice for cool drinks.
About as simple as it gets.
Everything is always wet.
Need for ice every couple days.
Clogs in drain are a real pain.
No ability to keep frozen food frozen (short of dry ice, which wasn't practical for us).
Vertical stacking storage is a pain.

Beneteau: 12-volt air-cooled.
Loved it, low energy drain.
Small freezer compartment could keep a steak or two frozen and make small quantity of ice.
Never had a problem with it.
Not enough ice production for 4 people.
Freezer too small for serious frozen storage.
Vertical storage still a bit of a pain, but side racks made it easier.

Gemini: Propane.
Larger box.
Door allows access to everything quickly without digging.
Freezer good for our frozen needs and making ice.
Even living aboard full-time and using the grill and stove almost every night, we use one 20-lb propane bottle [about $18] once a month.
Very cold.
We have a multi-hull; I understand that propane fridges like to be level.
The heat generated goes to the cockpit, which is nice on a cool morning, but on a hot day it's not welcome.

Other notes:

Purists may scoff at a unit with a door, but our system is so cold, it's not an issue; we monitor it with an external thermometer and it really isn't an issue. If I used more than one bottle of propane a month, I might think otherwise. Ease of access makes all the difference.

We still have a cooler in the cockpit. With a family of four, it's not realistic to keep everything in the fridge. So we'll put in a couple of bags of ice (which we use for our drinks) all our drinks and any veggies that want to be cool but take up a lot of space (lettuce, broccoli, etc). If it was just my wife and I, we could easily live with just the fridge. I really like having a big old pile of ice around. A coke in a big tall glass of ice tastes FAR better to me than one straight out of the can. Same for lemonade. We have one of those Coleman "five day" coolers.

Keep in mind, our needs are those of coastal cruisers; we shop once a week and have a diet very much the same as we have on land. An "ideal" system for a coastal cruiser with a family is going to be different than the "ideal" system for a single-handed long-range cruiser.

New Adler-Barbour ColdMachine's using Danfoss parts emit RFI that affects Marine SSB and Ham radios. Mfr has solution.
But from Gene Gruender on IRBS live-aboard mailing list:
We didn't notice it too much on our SSB, but it was bad on our VHF, even if we weren't receiving or transmitting it had a hum. It was noticeable by the people who were receiving our VHF transmissions as well.

I put a large filter from Radio Shack that was intended for a stereo in a car on the VHF power input and it helped a lot. If I could have found one that would have handled the amperage I would have installed one on the SSB and Fridge as well, but we were in Key West about to leave for the Bahamas when we went looking for something to fix it and that was all we found. I don't recall now what all I did to convince myself that the interference was all coming in on the power line and not the antenna, but I do recall that I was pretty sure that was how it was affecting it and the filter verified it.

The Adler Barbour unit uses a crude inverter to run the compressor and I expect that it is just a noisy unit. I would expect that some shielding and a lot of filtering, both at the compressor and the unit being affected, would be the only cure.

By the way, ours is about 5 years old and I have no idea if it has Danfoss parts or not.

From Colin Foster on SSCA discussion boards:
Mr. Kollmann's earlier book was a wealth of information - the systems I built from it have given top-notch service now for about 10 years BUT. Don't purchase his water-cooled condensers. They are badly designed. All three that I purchased from him failed in the same way. Freon erosion from the entry tube cuts through the water jacket in about 2 to 3 years, allowing salt water into your freon system. ... Although he was very responsive and helpful while I was building the system and purchasing parts, he never responded to any of my reports about the condenser failures so beware.

From Mary Mcatee on The Live-Aboard List:
Have had lousy luck with our fullsized Norcold fridge. It is AC/DC autoswitch and I can tell you it sucks a battery down right now. We also had to totally replace the fridge when it was three years old (total compressor failure).
From Michael Rich on The Live-Aboard List:
I also had one of the full size nevercolds and it proved true to its name. The compressor was replaced three times and you know how expensive that can be. I ended up cutting the whole unit out with a saws-all and putting it in the dumpster because the boat cabinetry was built around it. This is one company I will never support again for refrigeration. Now I'm getting along well with a 110 VAC 3.3 approx cube fridge running off a separate inverter and a couple 4d's.
From Bob Young on The Live-Aboard List:
I just want to put in a word here in defense of Norcold. I installed one of their icebox conversion units on my last boat, and it ran 24/7 flawlessly for 5 years, and was still running when I sold the boat. I installed another one a friend's boat, and his experience was identical. I chose this brand because they have a lower power consumption and are much cheaper than Adler Barbour. At that time - about 1994 - the whole kit cost about $400.

Another advantage was the ability to run on both AC and DC, and switching automatically to AC when available. Actually, as previously stated here, the compressor runs on 24 VAC, which is provided from 110 VAC via a transformer and from 12VDC via an inverter and another winding in the transformer. The AC powered hermetically sealed design is probably the most reliable and maintenance-free type of compressor - this type is used in almost all domestic refrigerators and air-conditioners, and these devices often run 20 years or more without maintenance.

From Roger Mummah on the SailNet liveaboard-list 4/2003:
... We have a Grunert air- and water-cooled frig. It has worked flawlessly since I installed it in 1997.

I installed the water-cooling plumbing (new thru hull too), but the tiny sea water circulation pump makes too much noise. In my infinite wisdom I "hard mounted" it to fiberglass and it makes the boat roar. Powerful little pump.

That's one of my projects - to find a softer mount for it. ...

From Norm on The Live-Aboard List:
We have had excellent service from our Engel reefers (one as reefer, one as freezer) and they have our highest recommendation.

The one used as a freezer a few months ago began having problems starting. It would take several tries, running for only 3.5 seconds, before it would stay running.

[The distributor] replaced the machinery on the spot and the trouble disappeared.

We highly recommend the Engel. Our two units, both MT-45's, have been in operation for almost two years and we are very happy with them. The one used as a freezer (simply the coldest setting on the thermostat) keeps ice cream hard and makes ice cubes daily.

[later] ...

My Engel reefer used about 1.8 amps when I measured it some time ago.

The unit set to 0 deg F for use as a freezer pulls 2.6 amps. The freezer runs almost all the time so it might use as much as 60 amp/hours per day.

I guess the reefer uses about 20-25 amp/hours per day depending on the ambient temp.

I recommend the Engel, but contact the Engel distributor, for the latest in his line of portable reefer units, the state-of-the-art is in rapid development.

My main complaint about the Engel is the insulation. The series is designed to operate having an unlimited supply of electricity and is not as efficient as would be a unit with more or higher R-value insulation. Perhaps there is available a unit with more insulation that would fit your space. I thought of adding insulation to the Engels, even have it on board, but never found the round-tuit needed ...

The more insulation, the less electricity required, but the less room for food available for a given outside dimension. There are some vacuum panels that claim high R values that might be of value on a portable unit.

Within the next year, I may be replacing my Engel's with SunDanzer units which have 4.3" of insulation. They have 4x the interior space and take about the same amp/hours.


> the Engel can sit out in the hot summer sun with a temp
> (I measured it) of 125 deg F on its 'skin' and be
> zero deg. F inside with only 18 AH per day consumption.

I find the above statement hard to believe. 18 AH/day is 9 hours of operation out of 24 for a duty cycle of .36. It was on 36% of the time. Our Engel freezer, set to zero F, in the galley, used more like 50 AH/day and operated almost continuously in the summer.

I have used two Engel units for about two years, one as a refrigerator and one as a freezer. They pulled about two amps each when running. These were the 1.4 Cu/Ft units.

Early on I put an hour meter on one of them and made some guesses as to AH used. My final conclusion was that both together used about 75 AH in the summertime with ambient temps around 90F, somewhat less in winter with the ambients around 70F.

Their main drawback for us was that the insulation is around one inch and they were small, 1.4 Cu/Ft. They seem to be designed for use in a motor vehicle with a plentiful supply of 12 to 24 volts where portability was important.

However, they were well designed by a German, solidly built by the Japs, and worked very well even considering the AH usage for their size. The were also self-contained, trouble-free, and sold quickly when I moved up to my present Sundanzer units.

Our current units, called "Sundanzer" are made in Sweden by Electrolux specifically for off-grid use. They are around 5 Cu/Ft, pull about 2 amps, and use the same amount of AH because the insulation is 4.3 inches thick.

We love them but they are a bit large for most boats.

[later] ...

I have aboard a Sundanzer refrigerator and a Sundanzer freezer. They are made by Electrolux for folks who live with solar panels, as I do.

They are 6 cu/ft, air-cooled, self-contained, 12 volt, 25 watt, quiet, maintenance-free and wonderful.
From David on The Live-Aboard List:
Used the BIG Engel before and agree not enough insulation thickness or quality - we are in the BVI's sheesh even the water temp is over 90 deg.

I would much rather have something 'good' than an almost works option ...

From Jeff Barfett on The Live-Aboard List:
The Engels use some type of 'sling' type compressor that doesn't eat much power. I've had my 45 running 24x7x6years - no issues, the 45 uses about 2 amps. I've thought many times about getting another, lose the case and install into the two deep iceboxes, but take a nap and forget the idea as I have a large investment into SeaFrost holding plates, which by the way make a lot of very large ice cubes!

Just a side note that Engel did at one time offer up the 45 without the case as a 'built-in'. Don't see that it is offered any longer.
From Norm on The Live-Aboard List:
The Engel guts are removable as a unit without disconnecting any tubing.

From Richard Kollmann on RParts Technical Forum 2/2004:
Re: a 12-volt freezer:

The Engel 45 is a top-loading cabinet unit that is AC/DC and will provide the results you are looking for. This unit is air-cooled and adds about one half the amount of heat to the cabin as a crew member would. The idea of water-cooling is not practical or reliable enough for these small units.

I tested most of the new small 12-volt refrigeration systems for my new book and here are a few of the Engel 45 performance figures: These tests were performed in a 24-hour controlled environment 78.5° to 81.2° F. in still air except for the unit's cooling fan.
As a day-use cooler at 46.9° F. the average amp per hour was 0.45 amps.
As a high temp. Freezer at 21.2° F. the average amp per hour was 2.78 amps.
As a low temp. Freezer at -2.65° F. the average amp per hour was 2.56 amps.

The amazing figure in the test as a low temp. unit, with a 82° drop in temperature the outside cabinet skin temperature only dropped five degrees.
From Forum Tech on RParts Technical Forum:
As for the energy consumption you are giving, something isn't right here. Based on the number you give, and the way you state the test results, for the "average" refrigerator draw (.46 amps/hr.) I assume that you are taking the total energy consumed and dividing it by the test hours. Indeed, to do anything else would be useless for comparing to other systems as it is the total amount of energy used per day that matters.)

On the other hand, if your "average" does include the total run time then you need to double-check your test setup as there is no way that the system has lower total energy consumption at -2.65F than it does at 21.2F.
From Richard Kollmann on RParts Technical Forum:
The tests I have done on the Engel 45 are correct and they did prove the value of this inexpensive Refrigerator/Freezer. The reason the Engel performs so well is it is not a conversion unit, it has a wrap around evaporator, and the components are a matched set for the boxes heat load.

I ran a series of tests using a cycle counter, Data logger, computer print outs to plot performance along with an amp-hour meter. If you check my figures against the manufacturer's figures you may find their figures are better than mine.

Average amperes per hour times 24 equals the daily amp-hours consumed.
Cooler 0.45 x 24 = 10.8,
Refrigerator 0.91 x 24 = 21.8,
High Temp. Freezer 1.47 x 24 = 35.3 and a
Low temp. Freezer 2.56 x 24 = 57.5.

Of the two hundred tests I have conducted in the last three years for my book, the ones I did on the Engel 45 were the most impressive. But like most items sold in the marine environment this unit is not salt water proof.
From Forum Tech on RParts Technical Forum:
Not to belabor the point, but your previous post claimed 2.78 amp/hr average for the high temp freezer and 2.56 for the low temp freezer. Also, what is the interior volume of this box?
From Richard Kollmann on RParts Technical Forum:
The 24 hour High Temp Freezer Test on the Engel logged 35.3 amp-hrs maintaining the box temperature at 21.2° F. and divided by 24 equals 1.4708 average amps used per hour. Bill was looking for a two cu. ft. freezer the Engel 45 quart is 1.5 cu. ft. Engel does have a larger 65 Quart model.

These Swing compressors have produced a better COP on smaller boxes than the larger Danfoss DB2.5. The newer BD 35 at its slow speed should approach the same COP of the 60 watt Swing compressor.

Just when we think the new variable speed Danfoss compressors are the answer to all of our small 12 volt conversion designs, I hear of a more efficient variable speed Rotary compressor to be available soon. The new Rotary is reported to have 30% higher capacity than the BD 50 and a 20% better COP rating.

If you need more information on these units go to Engel's web site. There are several other manufacturers of these self-contained units. If you recommend them as a low temperature freezer make sure the box insulation and the system is rated for the low temperatures.
From Richard Kollmann on Cruising World message board:
The tests I have been running are controlled in a static air area with a stable temperature control. I monitor each test with temperature data-loggers, cycle meter amp-hr meter, run time hour meter and total time clock. Three years ago I started evaluating some of the wild claims about boat refrigeration and box insulation. To do this I obtained many of the various manufacturer's systems without their knowledge. After hundreds of tests, I documented the results of what works and what doesn't.

In my new book on 12/24 volt refrigeration, there are a series of six evaluations of the Engel 45 refrigerator. Each test compares ambient air temperature against exterior box temperature to evaluate insulation quality. As to your question about condenser cooling, any air-cooled unit must draw air from one area and dispose of the super heated air to another area to be efficient. The refrigerator test listed with a power consumption of 16.1 amp-hrs per day was conducted with an ambient temperature of 81°F. and an average exterior box temperature of 77° F. As to using it for a freezer with the thermostat set to full cold, it should outperform any unit that does not have a full wraparound four-sided evaporator. As a low-temperature freezer with an average ambient temperature of 80.4°F, it consumed 57.5 amp-hrs per day running on 12 volts, with an exterior box temperature of 74.5°. With a delta-T of 6 degrees this is within an expectable range to prevent condensation on box exterior.

There are several manufacturers of these portable boxes and I am not saying it is the best. What I am saying is that it does exactly what Engel says it will do. There is no question that if any of these cabinet units were placed in the sun their efficiency would be affected. With seasonal changes or when cruising areas change, the seawater temperature has a major affect on cabin air temperature and refrigerator performance. I use a simple formula to adjust power consumption of air-cooled units located below deck. As the seawater temperature increases, add 4% for each degree F. of temperature rise or reduce the power consumed by 2% for each degree of temperature drop. An example of the same boat's power consumption in California could be 50% of what the same boat would use in July in Southern Florida.

From Charles Badoian on World-Cruising mailing list:
Using an Engel 35 right now and keeping it up with a 32-watt flexible solar panel on the fridge setting using a 27 AGM battery. Only issue I have with the device is that it rusts as we keep it under the companionway on our Corair 31 trimaran and can not prevent water from splashing it. The rust is on the outside so not really an issue.

Note we have minimal other power use but if we did I am sure the panel would not keep up.

Overall I am very happy with the unit and would highly recommend.

From "Angel Louise": bought a Waeco, and now wish they had bought an Engel instead.

From Steve Weinstein on The Live-Aboard List:
I almost went with the Frigoboat but was reluctant to cut another below-the-waterline hole in the hull. I discovered the Isotherm unit which, in most respects, is identical (including price) to the Frigoboat. The big difference is I didn't have to cut another hole. The Isotherm unit is also keel-cooled but they developed a "cooler" incorporated into a thru-hull which replaces an existing through hull. I already had a thru-hull for the galley sink drain (2" or 2-1/2" I think) and the Isotherm is a direct replacement. With the Isotherm thru-hull the sink still drains through that fitting but the heat from the refrigerator system is still dissipated.
From John Sexton on The Live-Aboard List:
This is the system I use. I am very satisfied with it. You do have to increase the size of the hole in the hull though.
From Steve Weinstein on The Live-Aboard List:
John, it probably varies boat to boat. In my case it was a direct replacement. The only modification was that I had to cut the drain hose to shorten it. The Isotherm thru-hull is a few inches taller than the standard one.

From Tom Henderson of AvXcel, manufacturers of the TropiKool, on The Live-Aboard List:
The TropiKool 40 runs continuously; it does not cycle on/off as do conventional compressors. The Electronic Control Module modulates cooler performance to match the current heat load of the box being cooled. This modulation is in the form of shortening the piston stroke of the cooler; this then decreases the power required during low load periods.

Under the Cruising World benchmark criteria, the TropiKool 40 consumes 18 amp/hrs per 24 hour period. Power consumption at 95 degrees F ambient, a R28 value, and a 40 degree F box temperature, the TropiKool 40 consumes 36 amp/hrs per 24 hour period. The continuous operation of the cooler enables the system to maintain a constant box temperature of +/- 3 degrees of selected temperature.


From Woodie Young in Benner Bay:
Never, ever turn off your refrigerator/freezer. Even leave it running when you're off the boat for weeks or months. If you turn it off when leaving, the system will heat up to closed-up-boat temperature, 100°F or more, greatly increasing refrigerant pressure and making a leak more likely, or increasing the speed of a leak.


From article by Glyn Johnson in 4/2007 issue of Caribbean Compass magazine:
"... the box is the most important part of the system. No matter whose refrigeration system you use, it will not work efficiently unless you have a good box, and if you have a really good box then virtually any manufacturer's system will do."

Summarized from article in DIY Boat Owner magazine, issue 1999 #4:
  • Many boat ice-boxes have water-saturated insulation; drill a test hole to find out. If wet, the entire box should be replaced.

  • All foam (except extruded polystyrene) should have a 1/2" air gap outside it, and should have no moisture barrier on the outside.

  • Don't put an extra cushion/insulation on top of the lid, unless you include an air gap.

  • Want refrigerator space about 2x size of freezer.

  • Want separate refrigerator and freezer spaces, not a combined "spillover" box (temperature is erratic and system is less efficient).

  • Sponge or pump excess water out of box, don't drain it into a sump or the bilge.

  • Gasket on door/lid is very important.

How to test insulating quality of your icebox,
summarized and derived from "This Old Boat" by Don Casey, with some additions from "The Boatowners Mechanical and Electrical Manual" by Nigel Calder:
  1. Get a scale, some block ice, and a thermometer.

  2. Pick a hot day, and close up the boat to get the interior hot (90-100 F).

  3. Measure temperature just outside the refrigerator/freezer box (ambient temperature) at various times during the following process.

  4. Have a typical assortment of food and drink in the box.

  5. Turn off the refrigerator/freezer and let it warm up (maybe to 50 degrees or so).

  6. Load ice into the refrigerator/freezer box, as high as possible.

  7. Wait until box cools down to 40 degrees.

  8. Weigh the ice (without any water) and put it back in the box.

  9. Wait some number of hours (at least 6; better is 24) and then weigh the ice (without any water) again.

Casey says a typical medium-size (9 cu ft) box with 2 inches of polyurethane foam insulation might melt 24 pounds in 24 hours (but he doesn't specify the ambient temperature).

Calder gives a Grunert-provided table of typical losses. It lists an 8 cu ft refrigerator box with 3" of insulation losing 3000 BTU (melting 21 pounds) in 24 hours (but it doesn't specify the ambient temperature).

Go on to calculate daily electrical requirements:

  1. Calculate daily BTUs needed to replace leakage:
    leakage BTUs == 144 * pounds melted * (24 / hours waited).

  2. Guess at the number of BTUs needed to cool newly-added warm food each day; maybe 1000 to 2000 BTUs. Maybe 150 BTUs per pound ?

  3. Guess at the number of times you'll open the door each day; maybe 10 to 20.

  4. Add BTUs needed to cool things, and replace door-opening leakage:
    total BTUs == leakage BTUs + (door-openings * 50) + warm-food BTUs.

  5. Calculate daily amp-hours (at 12 volts) needed to generate BTUs:
    current AHs == (total BTUs / 3.5) / 12.
    (Assume refrigerator removes 3.5 BTUs per watt.)

  6. Guess at the typical expected ambient temperature (just outside the refrigerator/freezer box) you'll have while cruising.

  7. Calculate multiplier between average measured ambient temperature and expected ambient temperature:
    multiplier == (expected - 40) / (current - 40).

  8. Calculate expected daily amp-hours (at 12 volts) needed:
    expected daily AHs == multiplier * current AHs.

For a freezer, Casey is unclear, but it seems you should:
  • Change the warm-food number from 150 BTUs per pound to 300.
  • Maybe lose 100 BTUs per door-opening, instead of 50.
  • Calculate the multiplier as:
    multiplier == (expected - 0) / (current - 40).

For a freezer, Calder says calculate the refrigerator number and double it.

"Refrigeration for Pleasureboats" by Nigel Calder shows how to calculate heat leakage while designing an icebox, from its surface area and insulation thickness. It assumes top-loading, closed-cell foam insulation, and no air leaks.

Paraphrased from "The Boatowners Mechanical and Electrical Manual" by Nigel Calder and "Refrigeration for Pleasureboats" by Nigel Calder:
  • Don't make box too large.
  • Don't put box in hot part of boat.
  • Want top-loading box.
  • Better if outside of box's cabinet is light-colored.
  • Want double (inner and outer) seals on the hatch.
  • Check for voids in the insulation.
  • Use closed-cell urethane foam (AKA rigid freon expanded polyurethane foam) insulation, not polystyrene or Styrofoam (expanded polystyrene).
  • Need a moisture barrier outside the insulation.
  • Good to have a heat-reflective barrier (aluminum foil) on the outside of the insulation.
  • Box must be airtight.
  • Box's interior liner must be strong enough to support heavy holding plates.
  • Design box so it's easy to remove the holding plate for repair.
  • Any paint on interior should have no lead content.

From John Branch on the Morgan mailing list:
I am currently building a refrigerator box for Maria and have spent a good deal of time researching insulation. The best that I could locate is Tuff-R by Celotex which has an insulation value of R-8 per inch of thickness. It's a polyisocyanurate foam and has slightly better insulating qualities than polyurethane foam. Comes in 4ft x 8ft x 1inch rigid sheets covered with foil and cost me twenty bucks per sheet. Initially I wanted to install six inches of thickness but after seeing how much interior space it took up, I settled on four inches.

Be sure to line the interior of your box with a plastic vapor barrier before you begin to install the insulation. Four to ten mils in thickness is fine and it will prevent moisture from forming and invading the insulation.

Vacuum insulation is available with a R-50 value at about $25 per square foot but it can fail for various reasons and then you're left with no insulation and a real problem.

I used an Adler Barbour Cold Machine on a previous boat for ten years with excellent results and for Mariah I've purchased an Adler Barbour Super Cold Machine which will handle up to a 15 cu ft box.

Instead of using a holding plate as the storage medium, I chose an evaporator and installed an 840 AH battery bank and let it be the storage medium. ...

From Bob Lucas on the Morgan mailing list:
At our Home Depot there is a section with insulation board. I have noticed that not all Home Depots carry the same boards. They tell me it has to do with local building requirements of the builders and remodelers. If you don't find it on the shelf in your area, you might inquire at the contractors desk.

From Dan / SunDial on the SailNet liveaboard-list:
We had an icebox on our previous boat. It was pretty large and we couldn't possibly keep it full of ice. So, here is what we did and it worked great for the year that we lived aboard:

1. We removed the teak shelf that held the ice and food off the bottom of the box.
2. We lined the entire icebox with Reflectix, taping the seams with the aluminum tape that they sell at Home Depot.
3. We took a cooler that would fit inside the icebox and double-wrapped the sides with Reflectix. The bottom had one layer of Reflectix, again taped to seal. The top was a single layer of Reflectix that would hinge with the lid. The cooler that we used was a Rubbermaid, I think. It was designed to allow 2 liter bottles stand up, so it was taller than it was square.

In this arrangement, we could put a 10 lb block of ice in the bottom of the cooler and it would keep as long as 5 days. This kept a small stock of drinks and food cold. Around the outside of the cooler but inside the icebox, we kept veggies, eggs, etc. This area was cool, but not cold.

All in all, I recommend Reflectix highly. It is very good insulation at a very good price.

From RParts Technical Forum:
> I found that I didn't have enough insulation by drilling
> 1/8" holes through the bulkheads and probing with a coat hanger.
> This worked well but I found I only have from 1-1/2 to 3"
> of foam. Also a lot of empty space on all 4 sides
> including 7" under the bottom.
> Since I can't get into that space, I intend to inject that
> spray foam (hardware store) to fill all the dead space, it
> will greatly increase the total volume of foam. Question:
> will it hurt that the foam will adhere to the 3/4 ply
> bulkheads and the fiberglass hull?

I hate to be the bearer of bad news, but you have a problem which is going to be much more difficult to correct than you are now thinking. You are correct in that you will indeed increase the amount of foam by filling in all the gaps with spray foam. What you are not considering is that in 6 months to a year (of constant refrigerator use) you will end up with a lot of very wet foam which will make you wish you hadn't touched it. It is not correct that "closed cell foam" does not absorb water. You must have some air circulation around the foam to prevent moisture buildup. The only way to really solve the problem is to rip out the box and completely rebuild it. If you are not up for that big of a task - don't do a thing.


Keep in mind that all of the information you are getting from various foam manufacturers applies to "normal" installations in a land-based application - not boats. The only conventional foam you are going to find that does not have a water absorption problem in boats is an extruded (not expanded) polystyrene.

The moisture passes directly through the cell walls so the fact that a foam is "closed cell" does not solve the problem.


Without having any clue about water absorption and condensation, I did what you want to do on an older Hunter.

Bought cans of foam at Home Depot and squirted foam into neatly spaced holes, thinking that when it came out of an adjacent hole, I could move to the next empty hole.

Did so.

Then to my horror, watched interior of box bulge inward up to 1 1/2" in the middle of the sides and on the bottom. Shelves would no longer fit, etc. Had to drill more, bigger holes in woodwork surrounding box and slowly dig out extra foam. Even so, each summer, the foam expanded a bit more. What a pain! Box got colder, though.

Caution! Caution.


Be careful when using expandable foams. Read the can for the expansion ratio and compare this to the volume you intend to use it in. These foams can and will exert a large amount of pressure if there is nowhere to go!

From article by G. Kevin Alston in issue 1999 #4 of DIY Boat Owner magazine:
  • Analyze your goals and your current setup before deciding whether to add insulation, build a complete new box, and/or replacing the mechanics of the system.

  • In the tropics, a refrigerator gets about 50% of its heat-load through the box insulation; a freezer gets about 80% of its heat-load that way. In a poorly-insulated freezer box, the food farthest from the evaporator and closest to the outside of the box may never freeze, no matter how good the mechanics are.

  • All foam (except extruded polystyrene) must be installed with at least a 1/2" air-gap between foam and any surrounding cabinetry.

  • Moisture barriers only serve to trap moisture after it gets through the inevitable pinholes.

  • Closed-cell urethane foam does absorb water; it just does it more slowly than open-cell foam.

  • Don't make the box interior liner out of stainless steel: it's hard to make watertight.

  • Radiant-heat barriers (alumninum-covered space blankets or bubble-wrap) are not very useful, since most heat transfer into a box is via conduction instead of radiation.

  • There is not much difference between a front-opening box and a top-opening box, because the extra air transferred with a front-opener contains very little heat. And a top-opener may be open longer because it's harder to find something that is buried.

  • Small gaps in the insulation do not cause catastrophic heat transfer, unless there's air-movement through the gap.

  • Under humid conditions, even the best-insulated boxes sweat when topped with a seat-cushion or similar insulator. This is not a sign of poor box insulation. Maintain an air-gap under the cushion.

  • If you find moisture in your insulation, you must rip it out and rebuild the box. Not only is the moisture killing refrigeration performance, but it may be causing hidden dry rot.

  • A cruising couple on a 40-foot boat probably would want a 5- or 6-cubic-foot refrigerator and a 3-cubic-foot freezer.

  • "Spillover" from freezer to refrigerator is not recommended. Temperatures will be erratic and efficiency low.

  • Don't install a drain unless you're going to use ice and no refrigeration.

  • Don't fit the box to the curve of the hull; make it rectangular.

  • Making a well-sealing door likely will be the hardest part of the box-building project.

  • Interior liner: consider having a plastic fabricator make one out of 3/8" polypropylene sheet.

From Kim on World-Cruising mailing list:
A refrigeration engineer told me that the air in the fridge or freezer doesn't really hold that much cold. When you open the door the impact has more to do with the entry of moist air, not the loss of "cold air". We see this especially in freezers and the ice buildup due to the moisture in the air that enters the freezer.


From N. Bruce Nelson on The Live-Aboard List:
> is a "holding plate" the equivalent of those blue plastic packs
> you can buy to freeze at home and then put in a cooler ?

Adding frozen blue packs would increase your "holdover". ...

The blue plastic packs contain an anti-freeze solution. Unlike ice, they will warm up as they thaw. Most American-manufactured holding plates do the same. Some holding plates behave like ice and stay the same temperature as they freeze and thaw. This is desirable because it results in more stable box temperatures and results in the food lasting longer.

From Stan on The Live-Aboard List:
If you're water-cooled, you need a brushless pump or you'll be replacing pumps all the time.


A standard 12V pump will die in a couple of months. I know, I went through several of them. The solution is a March model 893-09. It is a brushless, submersible (I don't submerge it) impeller-type, NOT self-priming, so it has to go below the waterline. My pump has been on my reefer for 3 1/2 years, only shut down to defrost. I think the cost is around $150, but it's money well spent.
From Colin Foster on The Live-Aboard List:
DO NOT purchase the March pumps that have the water-cooling feature that is fed by the pump itself. It is guaranteed to self-destruct. There are a couple of pipes that divert a little of the water through the pump motor for cooling. A really dumb idea. The first time the inlet or strainer is blocked up, the water flow stops and the pump motor cooling stops allowing the pump to overheat and self-destruct.

From Desmond Smith on The Live-Aboard List:
... use flare fittings instead of compression fittings. Either fitting will work. The downside of the compression fittings is that they can only be tightened a fixed amount until the ferrule gets crimped onto the tube. The flare fittings can be tightened until the threads get stripped off.

From RParts Technical Forum:
> I have a 15-year-old Iceberg water-cooled freezer/refrigerator
> system in my Swan. It appears to power two large holding plates
> in the freezer box (approximately 18 cu ft), this shares a common
> wall and cools the fridge box (approximately 12 cu ft) as
> well as a smaller holding plate for a small day fridge
> (approximately 8 cu ft).
> The compressor and evaporator finally gave up the ghost
> this season and I am attempting to figure out how to
> upgrade the entire system since it also used the older
> and now banned CFC's. What complicated this system is
> that the compressor could either run off a clutched pulley
> when the engine was going or off an electric motor when
> my 230v 50hz generator ran. So where do I begin?
> I obviously need all the upstream stuff to drive the
> holding plates, but how do I figure out what and how much HP?

The holding plates are your evaporators.

I would recommend trying to replace the condenser and compressor, and keep your system as close to its designed specs as possible. The condenser is easier to chose than the compressor. To be of more help I need a better description of the compressor (and condenser). Obviously, it is belt driven, but can you find any labels or tags that may have some manufacturer info on it? Or a physical description, including shape, size, material (aluminum, steel), and pulley sizes (both engine and compressor). Can you tell how many cylinders it has? Can you find any photos on our web site that look similar?

It isn't that the refrigerant is banned, it's that the refrigerants are very expensive, and harmful to the ozone. There is a good supply of recycled R12 available, but it is way expensive, and there are many interim "drop in" replacements that work very well.

The 12 volt compressors that we offer really are too small to do what you want to do, which is freeze two large freezer plates (probably plumbed in series). We do have compressors that will work, and condensers as well. Does your system have a water pump, and what does the condenser look like? What are the water connection sizes, and the refrigerant connection sizes?

From RParts Technical Forum:
Usually, refrigerator holding plates are around 24-26 degrees F, and freezer holding plates are around 0 degrees F. It is important to go several degrees below this temp to assure that the plates are as frozen as you can get them.

From RParts Technical Forum:
> Can buy refrig flare-fittings at hardware store ?

No, the flare connections for refrigeration are a different specification. They are "wrot" brass, and are stronger than the standard hardware ones. We sell fittings on our web site.

From Glenn Fagerlin on RParts Technical Forum:
Re: DIY Kit Experience - it works!

I assembled System 1 but made it water-cooled. I read the Calder book, tried to make sense of the calculations and decided, with some help from the Forum, to go for it. I would like to thank the Forum Tech, the Frigoboat guy on the forum and Alan, below, for their advise and help. I have a stainless box (3 cuft freezer + 4 cuft refer, front loading) built into my 45' Catamaran with a lot of nice maple cabinetry, so I decided to use it. Not the ideal but it was already there and someone spent a lot of $ on the box.

My timing was a bit off since I did most everything before the DIY systems, and instructions, appeared on the site. I am happy to report that the system works well, I learned a lot and saved some money in the process. Having read through the new instruction manuals and compared the recommended way with the ways I used, I conclude there are a lot of ways that work. Fortunately, I did not make any serious mistakes. Below are some observations that might help others.

1. I used the Aeroquip Quick Disconnects. This allowed me to pressure-test (120 PSI and soap bubbles) to find leaks (one solder pinhole and a few flares that needed tightening), disconnect the pieces and take them to a service shop (Swedish Marine in SF Bay Area) for evacuation and charging. I then installed the system, connected the components and fired it up. The service guy, Alan, at Swedish Marine was one of the original founders of RParts! He offered many words of wisdom as we evacuated and charged my system.

2. I found the high-temp silver solder (brazing) to be easier to work with. It came with the flux on the rods and flowed much better for me. The danger I learned is that the fittings, like the RParts water cooled condenser, are brazed. I used a heat sink like a wet rag and sheet metal to deflect the heat from the torch. For the disconnects, I placed then in a bowl of water that covered the brass but left the copper uncovered. This way the end of the fitting was a couple of inches from the brass part that houses the rubber seals. Both worked well in that the plastic covering on the condenser was not affected and the discoloration on the disconnect only got halfway to the brass component of the disconnect.

3. I got a controller from Frigoboat for the BD50 that allows me to set the RPM from 2500 to 3500. Without the controller the compressor runs at 2500, I am told.

4. RParts is a web-based company. They do not hold your hand, answer questions or otherwise operate as a normal retail store. Hey, it's their business model and they can run the company the way they want. The Forum answers most questions from what I can see, but there is no one to call at RParts for help. They do get you the parts. I ordered the disconnects on the web at about 4 PM PDT with overninght delivery. At 9 AM the next morning the UPS truck was at my door. Great service.


From RParts Technical Forum:
I recommend pressure for leak testing. Vacuum testing isn't adequate, as vacuum pressure is less than 15 psig, and since it is sucking, it is quite possible the leak won't show up due to flux, grease or other debris plugging the hole during suction. An internal pressure of 150 psig supplied by nitrogen or CO2 will be much more effective for finding leaks. Soap bubbles are a good way of finding leaks, also watch for a pressure drop over several days. Some of the nitrogen or CO2 will go into solution with the oil, especially with a temperature drop, but this should be less than 3 to 5 psig per day, and will usually stabilize after the first day. If the temperature doesn't drop, the pressurizing gas won't go into solution as readily.

Dyes work well, but the system usually has to be run for a while before the dye will show up, and once the dye gets spread around, it can be hard to pinpoint the leak.

From DG on Cruising World message board:
We might be replacing our refrigeration system. Currently we have an older Grunert AR-50 DC system. We're starting to have some problems, and the guys at Cay Electronics are saying that we might just be better off replacing the system, as ours now is about 16 years old.

Cay has offered a reasonably decent deal just to swap out a new Grunert system for the old, which would keep the existing plates, etc. They are recommending, however, that we consider the SeaFrost BD system which is both air and water cooled, so it has a little bit more flexibility.
From Richard Kollmann on Cruising World message board:
Before you decide on a replacement refrigeration system, review the slide show on my web site. If you are to replace the compressor and keep the holding plates, make sure the compressor's output is equal to the old unit's capacity. If the tubing in the holding plate is 1/2 OD then a 1/2 HP is the minimum size replacement compressor. If the boat has a generator onboard or has spent time connected to shore power I would also replace the seawater condenser.

Sixteen years is not an old Grunert unless the boat was used as a full-time live-aboard or has water damage.

One comment, there is no such thing as a bulletproof boat refrigeration system. Grunert, Frigoboat, Crosby, Adler Barbour, Technautics, Sea Frost all have excellent systems. The trick is to select the correct system for your application. If part of the box is to be a freezer and you plan on cruising in the tropics, sixteen year old insulation will have lost much of its insulation qualities.


All closed-cell foam insulation breaks down over time, allowing the trapped insulating gas bubbles in the cells to escape and be replaced with air containing moisture. Polyurethane ridged and spray foams are the most common foams used in refrigerators. What percentage of an insulation's R value is lost in ten years or twenty years is not something manufacturers would want to make available. I have heard the figure of 15 years used as the service life of closed-cell foam insulation, but not from any credible study. From my own experience there is two inches of spray foam in my 21 year old boat's refrigerated box. I see no measurable change in heat gain over the years but this observation means very little . So I still feel it necessary to point out the fact that insulation does age and become less effective over time. It is also my opinion that only boats with low-temperature freezers cruising in the tropics, need to be overly concerned about replacing dry old refrigerator insulation. If insulation is found to be wet then it must be replaced.

From Mike Brasler on The Live-Aboard List:
I have always felt "ripped-off" by the prices that "marine" fridges command.

We bought a domestic underbar refrigerator. It is modern, made with the latest insulation material, clean, white, and has natty little racks that the women in my life love. It has a small freezer compartment up at the top, and all the goodies in the door too. It came with a 220v 90w motor, and a 220v 15 watt courtesy light. I got a Danfoss BD 35F compressor and a 15w 24v lightbulb, and fitted them. It starts about 3x per hour for 7-10 minutes and consumes 2.3 amps at 24v including the fan. The charge is 47 grams of R-134a.

It has a concealed condenser using the case for radiation. We are thrilled with it.

The fridge cost 149 Euros and at that price I can afford to replace it every 2 years for the next 8 and still won't be close to the ridiculous overpriced madness that we are subjected to.

Most of the compressors are made in the same place regardless of the brand name anyway.


From Roger L on The Live-Aboard List:
Best book I ever saw on troubleshooting R12 systems was a Sam's manual on Automotive Air Conditioning by K. L. McDonald. Doesn't cover anything marine, but the principles still hold.

From Richard Kollmann on RParts Technical Forum:
Moisture on the outside of a refrigerator indicates the insulation is poor. Moisture inside a refrigerator indicates moist air is infiltrating into the box. Any time an object is ten degrees colder than the air, moisture will collect on the object. Moisture always collects on the refrigerant evaporator and forms frost in a refrigerator. Trying to eliminating evaporator frost by insulating it is not the answer. Keeping moisture out of the box is the answer.

From Woodie Young in Benner Bay:
Air-cooled refrigerator:

Touch the cooling coil on back.
Where it comes out of compressor, and probably goes to the top of the coil, it should be almost too hot to touch.
At the other end, probably at the bottom, it may still be warm but you can touch it.
If you can touch it everywhere, needs Freon.

From RLK on Sailnet forums 7/2010:
Repair Or Replace Your 12 Volt Refrigeration ?

The pleasure boat refrigeration industry today is following the marine industry down the drain. We have seen refrigerator companies such as Grunert, Crosby, Adler Barbour, WAECO acquired by Demitic. Bay has stopped production of their boat refrigeration products. Northern Lights, a manufacturer of marine generator sets, has acquired Ft. Lauderdale-based Rich Beers Marine. What seems to be missing in this new market are the experienced people who provided technical after-market support in the past. The technical approach to small-boat refrigeration today is the same as when you buy a toaster: if it does not work after a short warranty it is expendable, so they advise you to buy a new one. The most recent loss to small-boat refrigeration support is the announcement from that they will discontinue operations of Refrigeration Parts Solution effective June 30, 2010.

If someone tells you to buy new 12-volt refrigeration because your compressor will not run, do not believe them. If your compressor runs but there is no cooling, then the solution is difficult and generally requires technical help. If your compressor fails to run, then anyone with basic technical knowledge can find and correct a failure to run of a Danfoss compressor.

If your refrigeration unit is over 10 years old and has a Danfoss BD 2 or BD2.5 or BD3 compressor, then it has the older, discontinued electronic 4-pin module. Troubleshooting Danfoss compressors with 4-pin modules consists of the following steps:

1. All of these compressors have a 4-pin module connector and their modules contain an external fuse. If this fuse is blown there are two reasons why: either power wires to module are reversed, or module has an internal failure.

2. Check to see that there is actually power at the refrigerator control module.

3. Place jumper wire across thermostat terminals on electronic module. If compressor still does not run go to next step.

4. Disconnect black fan wire from electronic module. If compressor runs, replace fan. If compressor still does not run after fan ground wire is disconnected, go to next step.

5. Run correct size and correct polarity jumper wires direct from a fully charged battery in order to bypass all boat's wiring. Voltmeter readings are of no value when looking for voltage spikes. If compressor still does not run, electronic module needs to be removed and tested on another unit. If there are no other units available to test your module on, I will test all 12 volt Danfoss control modules free except for BD80 compressor modules. Email me for shipping address and details: richard at


Small 12/24 volt boat refrigeration using Danfoss compressors manufactured after 1996 will have a BD 35 or BD 50 variable speed compressor with a troubleshooting computer chip built into their control module. This circuit makes it easier to find troubled area if compressor fails to run. If your unit does not have this $2 LED install one, as it could save you a lot of money later. Without the LED on these new units. troubleshooting will be the same as earlier 4-pin Danfoss BD compressors.

If installed, troubleshooting LED will flash only if electronic module sees a compressor problem. In each case problems of compressor's failures to run are identified by counting number of flashes of LED:
  • No LED flashes would indicate either thermostat is open or no power to module.
  • One LED flash and a 4-second pause indicates a boat wiring electrical resistance problem or low batteries. Because of modules sensitivity to milliseconds of a voltage spike, they cannot be detected by a voltmeter. Solution is to bypass boat's wiring until problem is located.
  • Two LED flashes indicates fan over-current cutout. If fan circuit on these variable speed compressors exceeds 1/2? amp, compressor startup will be aborted. This condition can be confirmed by disconnecting Black fan wire at module; if fan runs replace fan.
  • Three LED flashes indicate excessive torque is required to start compressor. This is commonly caused by turning compressor off and back on too quickly or too much refrigerant or poor condenser cooling. Most people jump to the conclusion that there is a mechanical rotor lock up inside compressor and this is a mistake on Danfoss BD compressors.
  • Four LED flashes indicate compressor motor not reaching sustained controlling speed above 1,850 rpm quick enough.

If someone has tampered with refrigerant by connecting gauges to a system, letting air in or adding too much refrigerant can cause either a Three or Four LED flashing signal. On water-cooled Danfoss condenser systems, three and four LED signals are common when seawater gets into refrigerant circuit.

If you have sources for fair-priced refrigeration parts and technical assistance. email them to me and I will list them on my web site. I found a source for 4-pin modules at a fraction of the price currently being asked and have been advised that refrigeration parts outlets such as United Refrigeration can special-order the newer Danfoss modules or compressors at a fair price.

From Kollman Marine's "Servicing small refrigeration units":
Finding very small refrigerant leaks on these small units is difficult. I have four different types of electronic leak detectors that their manufacturers claim can detect a refrigerant loss as little as one ounce per year. With years of experience I am able to locate leaks of more than one half ounce per week. So I do not recommend you purchase an electronic unit. Electronic testers can speed up the time it takes to find a serious leak by locating a contaminated area but pinpointing leaks and confirming them still requires a positive secondary confirming test such as soap and water. My books have always recommended to boaters that leak detection by using a 50/50% mixture of dish washing soap and water is the best way for them to find refrigeration leaks. The correct procedure is to use the tips of a one inch brush to beat the soap mixture into a suspected area until it resembles shaving cream then watch area for two to three minutes to see if bubbles begin to form. Very small leaks in suction side of system can only be located with refrigerator box warm and system off due to low pressure in this area when system is cold or running.
Can of stop-leak fluid I bought has red dye in it, to expose leaks.

Refrigeration tools: see Tools section of my Boat Maintenance page.


AKA "Adding refrigerant" or "adding Freon" or "adding coolant".

Conflicting info:
I've gotten so much conflicting info, from various places, that my head is spinning. Some conflicts:
  • Getting air into the system would be very bad, yet instructions on the recharge hose I bought say nothing about purging air out of the hose, and neither do many articles. Some articles do say "purge the hose".

    Okay, mine is a short hose, maybe 8 inches long, so maybe not much air in it. But my system holds very little coolant, only 2.9 oz total.

  • Some people say overcharging will "kill" the system, others say overcharge and then bleed off the excess fluid until the system runs.

  • From Kollman article:
    > Is it ok to use refrigerant with stop leak included?
    > I am in a remote location if Africa and I can't get
    > pure refrigerant without stop leak in it.

    Stop leak mixed in refrigerant reacts with air and moisture in the air (humidity) to form a solid, polysiloxane (epoxy) seal at site of a leak. Leak Stop also incorporates an active ingredient that rejuvenates the seals, gaskets and O-rings in the A/C system.

    I would not recommend adding refrigerant from a stop leak can, as it might add additional oil or sealant, either one of these may permanently block refrigerant capillary tube. There are two types of refrigerant stop leak solutions. One has an active ingredient that rejuvenates the seals, gaskets and O-rings and the other reacts with air and moisture to form a solid epoxy. Any solvent that softens seals may affect compressor motor winding insulation.

    But just about every small can I see in stores is stop-leak fluid, not pure refrigerant, and the one I bought says nothing about whether it contains "epoxy" (seems to say it stops leaks by rejuvenating seals, so maybe no epoxy). Is "pure R-134a" available only in bigger cans ?

  • A minor mystery: here in Grenada, none of the auto-parts stores I've checked sell refrigerant stuff, and all of the hardware stores I've checked do sell it. What's with that ?

  • I assumed all refrigeration Schrader valves were the same, but after buying a recharge hose, I'm told automotive and small-Danfoss use different sizes (sure enough, it's true). And the packaging for the hose I bought says nothing about what size valve it fits (it does say "auto air conditioner" in the title). Hose manufacturer says it's a 7/16" connector; and "all R-134a low side ports for automobiles are a standard 7/16". Nova Kool says my compressor has "standard 1/4 inch" connector on it, but I measure the OD at more like 3/8"; maybe they mean it's on a standard 1/4" OD copper tube.

  • Some people say "you absolutely must use gauges to get it right"; many articles say "don't use gauges". Maybe it's good to use a gauge built into the recharge hose, and bad to use gauges that require piercing the copper tubing to attach them ?

    From Kollman article: "Suction gauge reading are risky so I prefer not to have inexperienced people connect gauges to this small system. Most experienced technicians would also not want to connect high pressure gauge to a unit that holds less than 5 ounces of refrigerant."

    From Kollman article: "Capillary tube refrigerators are designed to be serviced by volume in weight and not pressure. When the refrigerant volume is correct there will be frost covering 95% of a standard evaporator and none on the return line outside of refrigerator. It is a mistake to connect pressure gauges to a system unless you are positive compressor is running and frost coverage is not as it should be."

  • Most articles say "add coolant on the suction side", but I have this nice "process port" on my Danfoss compressor; is it okay to add coolant there ? If not, what is the "process port" for ? None of the Danfoss literature explains this. And that port is the only one I can connect a hose to; the suction and discharge ports have soldered connections to the main tubing.

    OTA article says charge via process port. Label on top of my refrigerator has a system diagram that labels the stub-tube coming out of the compressor as "suction access". Aha ! Nielsen article also identifies stub tube as "on the suction side". From RickB on Trawler Forum: "The process port is sort of a suction port, it is used for filling the charge during the production process ... the guy on the production line connects the charge supply there rather than passing it through the compressor itself as you are going to do with it running."

  • Are there two ways of charging, liquid (can upside down) and gaseous (can upright), and is this related to the process/suction port issue, or maybe the stop-leak versus pure-R-134a issue ?

    OTA article says charge via process port, with can upright (gaseous); doesn't say if compressor is on or off. Instructions on stop-leak can I bought say use can upside down (liquid), with compressor running. Nielsen article says "it's important never to tip the can upside down while you're charging". Kollman article always says "Add refrigerant slowly with refrigerant can upright adding vapor only". From "The Boatowners Mechanical and Electrical Manual" by Nigel Calder: "Charge with gas; never invert the can (the exception being the blends, which need to be charged as a liquid - consult a technician)."

    Does the liquid method require compressor to be turned on, and gaseous method require compressor turned off ? But Kollman article uses gaseous method with compressor on.

Info from various places:
Lots of bad things happen with an overcharge. R-134a in a refrigerator should be below zero psi pressure, about zero to minus 4 would be about right. Technique is very important !!! If you let any air into the system bad things are going to happen. Because it runs below zero psi, you want to connect and disconnect your gauge lines with the unit off and after sitting a few minutes so that the pressure equalizes and comes above zero. The air has to be bled from the hoses also so that it does not enter the system. Also, I would expect the TOTAL charge to be in the area of 12oz or so of R-134a, that is not very much! Small refrigeration units like this have what is known as a critical charge, standard practice is to weigh it in from empty but topping off is done by experienced techs at times.

Small refrigeration units are the most difficult to work on.


One of the key elements of small unit servicing is patience. You have to introduce an amount of charge, then wait for things to stabilize. They don't react quickly like auto A/C or home A/C units.


If you hire a technician to work on your Danfoss-based system, the first question you should ask is "What's the amount of refrigerant in the system"? If he doesn't have a clue then he is not qualified to work on a Danfoss/capillary tube system. Just putting gauges on these small systems can alter the amount of refrigerant.


In the old days we used to put in a bit too much freon, R-12 back then, and then we would watch the frost line, when a unit is a little overcharged the suction line will frost up when the system is running. You don't want it frosting all the way back to the compressor. Sweating is ok but not frosting. Then we would let a bit of freon out and watch the frost line retreat back into the box. We kept this up a little at a time until the frost line just disappeared. Small fridges don't usually have accumulators and are quite sensitive to the exact charge. Also when you are done the freezer temp will be lower. When it is back working and the compressor has cycled off for a while check your line tap with bubble soap for leaks. They often are a source of leaks.


From a cruiser: don't worry about overcharging, the compressor will detect it and stop. Dump in the whole can, try to run the fridge. If it won't run, bleed out a little fluid and try again. Repeat until the fridge runs. [But a couple of articles say "don't release R-134a (CF3-CH2F) into the air, it's still bad for the environment, just not as bad as older coolants".]

From RickB on Trawler Forum:
You don't have gauges so don't worry about the pressure. It varies with temperature anyway and whether the unit is running and you have to know what it means which I suspect you may not so don't freak out about it despite all the expert advice flying around.

Before you tighten the connection to the suction fitting on the compressor, crack the valve on your little can just enough to hear a hiss at the compressor. That gets rid of the air in the tube between the two.

With the unit running, give it a quick shot of gas and feel the evaporator coil to see if it is getting colder. If it is give it another quick shot and let it run for a while. If the performance is good, disconnect and see how long it lasts. If nothing changes then you have other problems and can choose for yourself how to handle it.

From Ken:
The valve on the Danfoss is a Schrader-type screw-on fitting like a bicycle tire. The usual Freon-refill cans available on the islands may have a different automotive-style fitting which will not fit. If you introduce any air at all into the system, or too much Freon, you're going to kill it. Be careful OK? :-)

From Kollman Marine's "Servicing small refrigeration units": "It is more energy-efficient to be undercharged a little than overcharged."

From Kollman Marine's "Servicing small refrigeration units":
The box and complete system must be warm when adjusting refrigerant charge. [I assume this means: system has been shut down for a while, and so evaporator is above normal operating temp.]

Adjusting refrigerant charge when evaporator is cold may not produce good efficiency.

Don't connect refrigerant gauges [or recharge hose, I assume] when compressor is running, as it is a good way to contaminate refrigerant.

My experience 12/2013:
  1. Did a VHF call on the cruiser's net and lots of boats volunteered to lend me a hose or gauge set. Some I think had automotive equipment that probably wouldn't fit their refrigerator; they'd never tried to use it.

  2. I used a hose (not gauge set) and 3 oz can of stop-leak.

  3. Had to bend the stub-tube on "process port" of compressor a little, to get clearance for the hose to fit past the condensor pipes.

  4. Compressor off. Put hose onto can, put hose mostly onto connector on "process port", turned valve to pierce the can, turned valve back to open the can. Some leaking at the can, which I can't quite eliminate. Purged air out of the hose (probably losing some gas in the process), closed valve on can (leak stopped), then finished tightening connection to "process port".

  5. Started compressor, opened valve on can, rotated can from upside-down to upright, mostly kept it upside-down. Can hear and feel compressor sucking fluid out of the can when inverted, can hear difference when can is upright and gas is being sucked. After 15 seconds or so, closed valve on can. Left everything attached, just let system run for a while. Noticeable improvement almost immediately.

  6. Let system run overnight. Evaporator (in freezer compartment) getting quite cold, but not what I'd call "95% frosted".

  7. Next morning, I tried adding more refrigerant, but I think the can was nearly empty; it must have all gone in when inverted for 15 seconds or less. Took hose and can off, cleaned up. Can seems completely empty: no gas or fluid shot out when I removed the hose. But I have no idea how much of the 3 oz of gas actually made it into the system; could have lost 1/3 of it while purging the hose and closing/opening connections.

  8. A little later, I think I am getting "95% frosted" in the evaporator. Maybe earlier I checked it at a "warm" time in its cycle. I'm happy with the results.

  9. Checking over the next several days, refrigerator is running fine, and I don't see red dye leaking from anywhere. I suspect refrigerant leaked out over a period of 5 years or more, so I won't see that leak anyway.

Peter Nielsen's "Recharge Your Refrigerator"
On-Time Appliance's "Charging 134a Freon On Residential Refrigerators"
Kollman Marine's "Servicing small refrigeration units"
Kollman Marine's "Small 12/24 Volt System Performance"